Usually it's the 12 (although they didn't use numbers for sizes back then). I don't have my (borrowed) copy of Dreyfus's "Type specimen facsimiles" handy, and I don't know all the names of the sizes by heart, but somebody should chime in soon.

On the other hand, I'm not sure the Stempel version isn't a revival (I mean not based on a single actual font) as opposed to a "literal" recreation.

hhp

A "Little Canon" (“Kleine Kanon/Misal”) was twice as large as a "Saint Augustin". Most probably sizes differed widely from one foundry to another and it was very difficult to mix types produced by different foundries or even by a single foundry.

Wait, I just remembered I have these!
http://www.themicrofoundry.com/other/l&g_070.gif _
http://www.themicrofoundry.com/other/l&g_071.gif

hhp

I don't really get this comment

I’m not sure the Stempel version isn’t a revival (I mean not based on a single actual font) as opposed to a “literal” recreation.

Does that mean that Stempel Garamond is not based only on this speicmen sheet alone? Or do you mean the oposite.

Sorry but my brain freezes up when reading dubble negatives in one sentence ;-)

thanks for your great help
Seb

Yeah, after I wrote that I suspected it might be too
labyrinthine... but then I got lazy and left it as is. :-)

I meant that Stempel Garamond might not be a "literal" recreation of a given size of the original (I mean even if they ironed out -presumed- printing artefacts, which historically almost nobody has ever left in, for any revival effort). It could be (and probably is) that to some extent, but quite possibly not enough for you to be looking too closely at a single size of the original. Note here that you can't ever fully trust the pedigree that might have officially been claimed at the time of its release - like how even to this day some Jannon revivals still try to allude to being a Garamond, all these decades after Warde's article. My all-time favorite is this classic bit of Copperfield-grade obfuscation:
http://www.myfonts.com/fonts/linotype/simoncini-garamond/

Somebody around here should know about the Stempel for sure.

hhp

Based primarily on the counter of the 'a' and 'e', I'm going to say that the best match is the "Romain Cicero", which according to the Fournier Manuel Typographique is approximately 12pt. The next size up, the St. Augustin, has distinctly smaller counters as appropriate for display.

The Linotype material I have boasts of Stempel's possession of an original copy of the Berner sheet, and to my eyes, it's a fairly literal revival. Since I'm looking at the Linotype version, some of the differences (short descenders, fairly wide 'm') may be due to that adaptation rather than what Stempel did.

In any case, the Stempel version does honestly seem to be based on the Garamond model, rather than Jannon or Aldus or anyone else. Also, looking at (my reproduction of) the Berner sheet, I'm struck by how beautifully the optical scaling has been done.

Of course, now that Garamond Premier Pro is out, pretty much all the fun has been taken out of Garamond revivals. I was just looking at my Le Be from a few years back, and am not sure I'm ever going to finish it.

I read over the previous thread, and can highly recommend the "Type Specimen Facsimiles" edition. The reproduction there, while not perfect, is vastly better than the GIF attached to your thread. I also have high-res scans of that facsimile, with 200 dpi versions here and the 2400 dpi original scans at the TeX archive.

The best Stempel Garamond source I've found so far in my library is Hermann Zapf's Manuale Typographicum. According to the book, "HZ and his design philosophy," "the type specimen of Konrad Berner of 1592 was the source for the redesign of the Stempel Garamond." Examining it more closely, I see that Stempel was in fact responsible for the shortening of the descenders (they were German, after all). They did a creditable job, but something has definitely been lost in the elegance of the 'g'.

I'm fairly sure now that they didn't just trace over a particular size, because I see elements from different sizes mixed. Overall, the Garamond is remarkably consistent across the size range, but there are some systematic differences such as 'a' and 'e' counter size, as well as quirks such as the height of the upper left serif in the 'p'. In the Cicero and Galliard, it's just a touch high (a tendency which is greatly exaggerated in the Jannon branch of Garamond-named fonts), but in the Petit Canon it looks just in line. The Stempel follows the large sizes here. (Incidentally, the Stempel 'p' is also considerably wider than the Berner sheet. I think it was traditional for the bowl of the 'p' to be a wee bit smaller than the 'o' norm, something that gave me an inordinate amount of trouble when digitizing the Le Be Book. This quirk is completely normalized out of the Stempel. The Stempel has a more pronounced pen angle, as well, suggesting that this letter was redrawn from scratch).

The 'a' tells a different story. There's no way the Stempel could have been drawn from the large sizes, which show a very smooth and refined bowl, and a very light touch on the terminal. Stempel is characteristically quirky in both regards, and also has a general lean to the left (a wide top) completely absent in the Canons. The heavier terminal weight and top width vary unsystematically across the size range, but the Cicero is certainly not far off.

In other tiny details, Stempel doesn't follow the Garamond closely at all. For example, in most sizes the top right serif of the 'u' is much heavier than the top left, but in Stempel they're only subtly different. Also, the Stempel 'i' dot is big and bold, much more subtle in the Garamonds. The Stempel 'M' is much more symmetrical left-right; Garamond in all sizes has noticeably more white under the right side of the vee than the left. A lot of these changes count, I think, as regularizing the alphabet to match twentieth-century expectations.

Hope you find this useful. It was a pleasure to go back to the Berner sheet and study it closely again. Garamond does indeed deserve much of his fame.

Fantastic help.
Thanks to all of you.

Nice, Raph.

> all the fun has been taken out of Garamond revivals.

No - I have an idea.

> something has definitely been lost in the elegance of the ‘g’.

And something gained, seeing as how it's supposed to be a text face.
A good text "g" needs to have a certain inelegance.

hhp

And something gained, seeing as how it’s supposed to be a text face.
A good text “g” needs to have a certain inelegance.

How about we split up the font design work between us? You do the inelegant text fonts, I'll confine myself to elegant ones.

Incunabular romans generally have very long descenders (small x-heights) by today's standards, but I have found a couple of very old examples that could almost pass for '70s ITC designs. One is the 1689 Plininae exercitationes by J. Vande Water, Utrecht (reproduced as plate 184 in Day and Morison's Typographic Book), and another is the Groote Garmont Romeyn in the specimen sheet of Jan Roman, Amsterdam, c. 1762 (reproduced in the Type Specimen Facsimiles edition, p. 13). A couple of things to look at next time you head for the library.

You get the pretty ones, I get the interesting ones - sure.
But could we maybe switch on nights out?

> very long descenders (small x-heights)

Wait a second - those two are not the same. My support for short(er) descenders in text fonts is not tied to ideas of good x-height versus extenders proportions. It's based on letter frequencies.

> A couple of things to look at

Thanks!

hhp