Calypso PF free download

Excoffon’s Calypso now free for download!

During research for my book Letter Fountain many typefaces passed my eye. Most of them designed with meticulous precision and based on the knowledge of readability and technical perfection achieved in hundreds of years. But sometimes fascination strikes when a designer leaves the common path. Roger Excoffon was one of them and for me one of the most striking typefaces was Calypso, a typeface that has a terrible readability, and you can even ask yourself why it is made in the first place. Marcel Olive, owner of Fonderie Olive saw Excoffon experimenting with an enlarged print of a half-tone screen at Olive studio. He was rolling it up and looked through it like a kaleidoscope. A metal type with half-tone dots was not done before and a technical challenge to achieve. Marcel Olive saw the chance to profile the technical capabilities of his foundry and earn a worldwide reputation and gave Excoffon permission to execute the design proposal.

After establishing the angle and size of the dots by Olive Studio each character was drawn dot by dot using a pair of compasses. According to José Mendoza y Almeida, who lead the team at the studio, Excoffon made sketches of the outlines of each character and in the studio shading was added by airbrush. The airbrush shading was converted to a dot-screen that went from deep black to white. It was quite a challenge to transfer the drawings with a pantograph and to scale this complex drawings in different type sizes to the matrices. Then it had to be milled, retouched and casted in lead reproducing all the dots of the dot-screen. Calypso was cast in four sizes: 20, 24, 30 and 36 pt and had 26 capitals, a period, an apostrophe (used a lot in French), and a hyphen.

Most of the typefaces ever made have been digitized. Calypso was no exception. I found and downloaded Calypso Boy from Scootergraphics (Digitized by Marty Pfeiffer) and Calypso by Profonts (digitized by Ralph Michael Unger). Ralph Michael Unger has added numerals, a question mark, an exclamation mark, ligatures and a lot of other useful characters, making it a complete digital font. By comparing the capitals I saw that they where quite different and it seemed to me that they were based on the Calypso silkscreen-printed rub down Letraset version because the dots were not round like on the original drawings I had seen in several publications and advertising for this typeface.

Of course the original drawings were also not exactly the same as the metal type. As earlier written the punches that were cut by the Benton pantograph were retouched and because of that there were differences compared with the original drawings. So the final design had to be found in the actual cast type. I went looking for this type and found the site of D. Stempel GmbH that got the original matrices of D. Stempel AG and all the takeovers Stempel made during their existence. One of them was Fonderie Olive. I ordered a set of newly cast type from the original Olive matrices and found out that it was indeed quite different from the digital fonts that I bought. At that time Marjolein Koper was working as an intern at our design studio Polka Design and I asked her to digitize Calypso. The result was better than the fonts I bought but still I was not satisfied. After she came back to work at our studio on a steady base we photographed the metal type with a Micro Nikkor on a D800 to get the sharpest enlargement we could get. With this pictures Marjolein established the exact angle of the grid and we decided to begin again from scratch. Although it still is not an exact reproduction of the original metal type it has more detail and it can match almost the big reproductions seen in the first advertising in the French printers’ yearbook “Caractère Noël 1957” and recent publications with original drawings. You can download the result of our effort free on our site as TrueType font for Mac and Windows: http://www.letterfountain.com/extras_e.html
Please tale advantage of this utterly useless font in your work and send me some results!

Joep Pohlen, Polka Design/Fontana Publishers, April 2013
Steegstraat 12, NL-6041EA Roermond, The Netherlands
joep@polka.nl
www.polka.nlwww.letterfountain.comwww.crucialfuel.com

In the picture you see above left the original sample of the 'C' from the Calypso from 1960 in the yearbook Caractère Noël. Below right you see our version. The other two are existing digital fonts from 1997 and 2005. Above the 24 point metal type our version is based on. Of course together with other sources that show original drawings and enlarged type.

Further reading:
Sandra Chamaret, Julien Gineste and Sébastien Morlighem, “Roger Excoffon et la Fonderie Olive”, Ypsilon, Paris 2010 (French/ English).
Rault, David, Roger Excoffon, Le Gentleman de la typographie, Atelier Perrousseaux, Paris 2011 (French/ English).
And a recent article in Eye Magazine featuring an original Calypso ‘R’ drawing on the cover:
http://www.eyemagazine.com/feature/article/mr-mistral

Birdseeding's picture

Wow, those are some particularly crappy alternate digitizations. :o Nice job!

Queneau's picture

Lovely! I always was fascinated by this Typeface, thanks for the digitalisation.

polkawithfontana's picture

Today I received the yearbook Caractère Noël from 1957. In this yearbook the first advertising came out for the Calypso. Calypso was launched in 1958. The yearbook of course was published also in 1958 but the work shown in the book should be dated 1957. Enclosed two pictures from the advert, one with a transparant page in front. Probably the airbrushed and shaded 'C' was the original for this character on which the dot screen was based.

hrant's picture

Very impressive. Your dedication is commendable! And that photo of the metal version is amazing. BTW did you try scanning the original (maybe after inking the face) instead of photographing it? I would think the latter method would introduce parallax. And what was the effective resolution of the photos?

BTW, a side question: do you know whether Monotype (via Linotype) now owns the rights to Calypso?

hhp

polkawithfontana's picture

We did not scan it. I think that inking (or like they did it in earlier years blackening it with a candle) would not make a better image but I did not try it. We have corrected the pictures parallax and because we made a grid (look at our YouTube video) and the fact that this grid fitted on each character we presumed that we corrected the parallax. The Micro Nikkor we used is not a zoom lens so has not much deformation in the lenses so I think that it is quite accurate. The Nikon D800 has a 36 megapixel resolution and is by that almost a scanner (almost 100MB files). You have to remember that the drawings by Almeida where of course made by hand and so was also the following of the contours on the pantograph done by hand. So in that stage there is also some inaccuracy. I think also that their goal the mechanical look was and not a handmade look. We considered to scan it on our Epson V750 Pro which is also a top scanner but also works with lenses (deformation). My other consideration was that you can create shadow and a sharper difference between side and top of the metal letter with photography while a scanner does not create shadow because the light source is almost where the scan unit is.

I think that Monotype has officially the rights after Olive, Haas, Stempel, Linotype and Heidelberg. But the Profonts Calypso was also not published by Monotype (and the Calypso Boy was also a private project). When Monotype makes a better one I will take it of the website of course. The only goal is to have a proper and usable Calypso.

hrant's picture

Great explanations, thanks.

hhp

5star's picture

Thanks for the font, I'll use it somehow somewhere it has a nice depth to it.

n.

polkawithfontana's picture

Don't get me wrong about the legal rights of intellectual property. All the fonts we use in our commercial projects of Polka Design and Fontana Publishers are bought and licenses are paid. The making of the Calypso is a tribute to Excoffon and has to be seen like that. The license holders have had more than twenty years to publish a good digital font but decided, I think on commercial grounds, that it was not interesting enough to do so. On the other hand I think that the legacy of Excoffon and Almeida is important to the typographical community and for young students that have to be confronted with the mind set of designers like Excoffon.

hrant's picture

I'm certainly no fan of squatting myself. As long as people are warned.

hhp

polkawithfontana's picture

I now see that I forgot to mention the link to the YouTube video from 'The making of Calypso', Sorry for that! It makes clear how we worked on the pictures from the metal type with the grid.
http://www.youtube.com/watch?v=UTbwg7Sz4Xc

I corrected this link so now it should work!

PabloImpallari's picture

Joep: Awesome work! Congrats!

John Hudson's picture

Very well done. You've got some distortions to small curves in the small white spaces, and a few outline overlaps that will cause knockouts in some renderers. For a design like this, I would have favoured PostScript over TrueType, I think. But overall this is very effective, and certainly a lot better than the other digitisations.

polkawithfontana's picture

It is a complicated discussion about licenses and intellectual property. There are always two parties involved. One is the holder of the rights and one is the maker of the work and his or her heritage. Just like paintings and other art the common opinion among collectors and museums is that you may buy a work of Van Gogh for several millions but you never own it. You pass it through to the next generation and you even have the obligation to take good care of it. Some time ago a Japanese businessman had put in his will that he wanted to take his painting of Van Gogh with him in his grave. Legally it was not possible to do so because of the reason stated above. It was public heritage. I think that Excoffons work belongs to our graphic heritage and that it has to be excessible for study and use. Preferrable by the rights owner of course.

You are right John. There are still some design flaws in our Calypso. Anchor points can still be reduced and there are some overlaps. We will make the font better in the coming time but I think that this still is a big step from what we had. We tested it and we had no problems with it.

PabloImpallari's picture

This macro can help you to find and remove overlaps:
http://typophile.com/node/84799#comment-476051

polkawithfontana's picture

Thankyou Pablo for your input to remove overlaps.

About the copyright of the typefaces from Olive, I found info in the book 'Roger Excoffon et la Fonderie Olive'. On page 67 the following is mentioned: Haas did not obtain the copyright for the typefaces but was granted the permission to cast selected typefaces. Therefore Olive sold the casting machines, pantographs, matrices and 100 tons of typefaces to Haas. Until his death in 1990, Marcel Olive retained control over the rights of the typefaces. He granted several foundries the non-exclusive reproduction of typefaces. Among them: Bobst, Compugraphic, Itek, Mecanorma, Letraset, Hewlett Packard, Linotype, Agfa, Adobe and Scangraphic. This information is said to be found in the archives of Marcel Olive (as mentioned in this publication).

hrant's picture

If so, Olive's heirs might still own the rights.

hhp

dberlow's picture

this is fabulous, I've been waiting for this so long I can't remember what I wanted it for. ;)

John: "...and a few outline overlaps that will cause knockouts in some renderers. For a design like this, I would have favoured PostScript over TrueType, I think."

I think, if a design has overlaps, the only format with overlapping contours legal in the spec, is TT. Removing overlaps in a design like this to assure correct PS output could introduce distortions in scaling.

polkawithfontana's picture

Yes, they might still be owners but on the internet you can not find anything about it. Though, interesting is also the statement on the site from Jean Francois Porchez from 2005. According to this page he acquired Fonderie Olive ...
http://typofonderie.com/gazette/post/porchez-typofonderie-acquire-fonderie-olive/

Albert Jan Pool's picture

bbg: ‘I think, if a design has overlaps, the only format with overlapping contours legal in the spec, is TT. Removing overlaps in a design like this to assure correct PS output could introduce distortions in scaling.’

Why scale down? You can do an OpenType font with PS outlines and leave the PPEM at 2048. FF DIN Round has a ppem size of 2000 units/em. During testing it in 2010 one of the PDF-Writers made a mistake because the programers had forgotten to look into the FontMatrix. They had simply assumed that TT-flavoured OTF is always 2048 and PS-flavoured OTF is 1000 units/em. Luckily they took notice of this and updated their PDF writer. AFAIK, FF DIN Round has never caused any trouble ever since. Adobe also produces Arabic OTFs with a ppem-size of 2048 units, so there are at least two manufacturers which took care that the pavement has been cleaned. So my advice is to stay with PS outlines, remove the overlaps and avoid scaling trouble by sticking with the ppem-size of 2048 units/em. I think that users who import your outlines in Illustrator will appreciate that too.

Albert Jan Pool's picture

polkawithfontana on 10 Apr 2013 — 12:44am:
‘According to this page he acquired Fonderie Olive ...’

To good to be true … you should have posted this 10 days ago :–)

dberlow's picture

"Why scale down?,

AJ, I did not suggest scaling the em in the development process. My comment was related to the scaling done during use, which is as likely to be down as up. So, go back and think about the removal of overlap when not required, and what that would do to hundreds of previously perfect circles, now conglomerated into point lumps, then scaled during use.:)

Albert Jan Pool's picture

bbg: ‘hundreds of previously perfect circles’

You are right, even when one would stay with ppem 2048, removing the overlaps would cause a lot of extra trouble. I have seen TT fonts with overlapping outlines (Ccedilla) causing incorrect (PS?) output though, so at least some testing and a closer look at rendering methods might be useful in order to get a clearer picture of the pros and cons.

hrant's picture

Joep, as we say in American: same difference. :-) I don't see how nothing being on the Internet and/or somebody else claiming ownership changes anything.

Luckily they took notice of this and updated their PDF writer.

What if they don't?
In the user's view the blame always goes to the font.

BTW is the assumption that a TT font must be 2048 rarer than the assumption that a PS font must be 1000?

hhp

Jens Kutilek's picture

What if they don't?
In the user's view the blame always goes to the font.

Ask the users of XeTeX, they have a 4-year-old filed bug about this issue: http://sourceforge.net/p/xetex/bugs/23/ – it doesn’t seem to be that urgent ;)

BTW is the assumption that a TT font must be 2048 rarer than the assumption that a PS font must be 1000?

From my anecdotal evidence, yes. TTF seem to work anywhere with any upm (allowed are 16—16384)*. For a typeface like Calypso I’d have gone for at least 4096, because you can see in the 1000 upm version the circles can’t be constructed very faithfully.

* There are some issues in Word on Windows with high upm + high coordinate values (very wide glyphs), the font smoothing is disabled in this combination.

Bert Vanderveen's picture

According to this page he acquired Fonderie Olive ...
http://typofonderie.com/gazette/post/porchez-typofonderie-acquire-fonder...

It WAS dated April 1… So…

polkawithfontana's picture

Just like Albert Jan Pool you are fully awake ...

polkawithfontana's picture

>Joep, as we say in American: same difference. :-) I don't see how nothing being on the Internet and/or somebody else claiming ownership changes anything.<

Hello Hrant, my answer to that I gave before: >I think that Excoffons work belongs to our graphic heritage and that it has to be excessible for study and use. Preferrable by the rights owner of course.<
As long as this is not the case our effort should serve the ones who wants to study and use it. However, I understand the point you are making ;-)

As suggested I tried to scan the metal type on a resolution of 4800 dpi to see if it is better than the pictures that I used made with the Micro Nikkor and Nikon D800. The scan is sharper but does not give more detail. I think because it is already an enlargement and I have reached the (lack of) sharpness of the casting.

hrant's picture

The one on the left is sharper, and probably usably so. Is it the scan?

I have reached the (lack of) sharpness of the casting.

I don't know - type metal molecules are pretty small. :-)
But of course there were/are coarser limits in the production process. What's the coarsest limit, the pantograph? BTW what point size was the metal type you used?

BTW, those close-ups make this project doubly fascinating. If I were you I would send presentation proposals here:
http://www.atypi.org/atypi-amsterdam-2013
http://www.typecon.com/archives/2603

hhp

polkawithfontana's picture

Hrant, the one on the left is from the Nikon D800. This is the 24 pt version of Calypso. I am considering buying the 36 pt version, the largest one ever casted. But that is a big investment.
Maybe you get some loss from punch to matrix and from matrix to letter. It could also be that you loose some sharpness when a lot of letters are made in the same matrix. But I think that Calypso is not cast in the same quantity as Times New Roman :-)
I will look at your links. Shortly I will get also a Letraset sheet with a 72 pt version of Calypso. I hope that I still can rub it down. I will make some pictures of it as soon as it arrives!

hrant's picture

Huh, the camera is better!

I highly doubt they only used punches for Calypso - there are too many things stacked against that possibility. They probably used some sort of pantograph (in fact maybe that constituted the Olive foundry's technical superiority that allowed the production of Calypso) and the limit there was the size of the cutter (which IIRC is on the order of 1/500 inches for the smallest one*). What's interesting here is that a pantograph's cutter is circular... a perfect fit for Calypso! No loss of "resolution" due to the size of the cutter, if you cut directly into the matrix. However I'm not sure how the "shearing" of the circles (by the swooping curves) would end up working in such an approach... Maybe there was another step (actually there might have had to be two, two back to "negative" after a "positive") before the matrix.

* They used bigger ones for clearing out large patches first (without breakage) then switched to a small one for details.

Related: http://typophile.com/node/1083?page=1
See for example the last image there.

Letraset: Do you have to rub it down to scan/photograph it? BTW another angle might be phototype - did Calypso ever have a photo version?

hhp

Mark Simonson's picture

The Letraset version should be an interesting comparison. They normally did the production artwork at several inches tall cut in Rubylith film.

Also, you wouldn't necessarily need to rub it down—just photograph or scan the reverse side of the sheet and flop the image.

kentlew's picture

What's the coarsest limit, the pantograph?

Several manufacturers were in the habit of handing out promotional souvenirs with the Lord’s Prayer engraved on a single 6-, 8-, or 12-pt em square, cut with a Benton punchcutter and cast as a type sort, as a demonstration of technical prowess. I’ve heard of these from ATF, Monotype, and Linotype. There may have been others.

hrant's picture

Kent, thanks for that reminder.

So I took some measurements in Photoshop and the smallest visible feature in that sort on Flickr (which I think is the closed counter of the "A") is about 0.7% of the Em. If the Em there is 4mm that comes out to just over 1/1,000 inches. So twice the resolution that Giampa states in that other thread. But still much coarser than Joep's scanner or camera.

Now, 4mm is about 11pt - presumable it was actually 12pt. If they did go down to 6pt that doubles the resolution. However we don't know if things like the counter of the "A" survived at 6pt. And 2000 dpi is still less than half the resolution Joep is working with.

Also, it's possible they used an especially thin cutter for that project - so thin as to break too often to be practical for real work.

hhp

polkawithfontana's picture

Here is the one from Haas. It is the Gutenberg song (Gutenberglied) engraved on a 12-pt em-square. I placed a match under it to compare the size. They called it 'Ein Wunder des Schriftgusses' (A miracle in typecasting). It has 357 characters. I believe that it was made for the 400 year existence of Haas in 1969. At that time they had taken over the casting activities of Deberny & Peignot, Olive, Berthold and Stempel.

polkawithfontana's picture

I don't know if Calypso was made for phototype. I doubt it. In the list of sold material to Haas when Olive stopped in 1979 only metal type equipment was listed. But already in 1974 an number of major foundries approached Olive to adapt types for phototypesetting. In 'Roger Excoffon et la Fonderie Olive' several typefaces are mentioned to be published for phototyepsetting, of course Antique Olive (a.o. by Monotype) but also Mistral (by Itek) and Banco.

hrant's picture

The Haas one looks comparable to me.

Something to repeat/clarify though: if you're cutting shapes with no tight corners (i.e. Calypso's circles) the size of the pantograph's cutter is irrelevant (as long as it fits inside).

hhp

5star's picture

polkawithfontana, that Gutenberglied thing is friggin awesome!

n.

Indra Kupferschmid's picture

Calypso was available at Photo-Lettering and VGC (and Castcraft) according to Hans Reichardt.

dberlow's picture

Of course it was! And it was made exactly the same way every other film or metal face was made at that time. The idea that they would triple or quadruple the costs by any deviation from the norm, is invented here for the first time.

Indra Kupferschmid's picture

Oh, and there is also a “line”-version of Calypso by Marty Bee as sighted by Florian in Berlin.

hrant's picture

David, if anybody's guesswork is wrong please explain how it's wrong, so we can learn. And -as Joep reminds us- the Olive foundry had a technical advantage over its competitors (that in fact it wanted to flaunt via Calypso) so not everybody could have been doing it "exactly the same way". BTW the point of that line of discussion came from Joep's "I have reached the (lack of) sharpness of the casting."

Indra: The line version is cool. Fewer points! (Although that's not relevant to film.)

hhp

Indra Kupferschmid's picture

The line-version is digital.

joeclark's picture

YouTube video is marked as private.

HVB's picture

Hard to tell which came first - but the Marty Bee 'Line version' is internally identical to Rafael Dinner's "Stiletto Silver" who also did a "Stiletto Black". Both can be seen here. - Herb

hrant's picture

Joe, this one (from the site) works:
http://www.youtube.com/watch?v=UTbwg7Sz4Xc

HVB: Indeed the "O" (and some other glyphs) in Stiletto Silver seems to have the same problem (although I guess it could be a feature) as the one in in Indra's link.

hhp

eliason's picture

Something to repeat/clarify though: if you're cutting shapes with no tight corners (i.e. Calypso's circles) the size of the pantograph's cutter is irrelevant (as long as it fits inside).

But look carefully: In every glyph of Calypso, at some point in the darkening, the circles start to overlap, producing negative squarish white space with quite sharp corners.

hrant's picture

Ah, you're right. So maybe the pantograph (assuming they did use it) was the "resolution bottleneck" after all.

A key thing to nail down would be exactly what Olive's technical superiority was.

hhp

dezcom's picture

Amazing amount of work for no compensation?

polkawithfontana's picture

Indra, thanks for pointing to a line version. Another attempt was made by Marty Pfeiffer and the name of his project is Belafonte. He made a digital shaded version of Calypso and a line version. Enclosed both pictures. The project can be downloaded on his site: http://www.scootergraphics.com/index.html (at least I did ...)

Syndicate content Syndicate content