I picked up a copy of “Origines de L’Imprimerie En France” (Arthur Christian, 1905) a couple weeks ago, and among its many delights was a font with the following attribution: Cette premi
I remembered that H
OK, there’s something in the IN specimen book (1990). From page 134: “Arthur Christian, directeur de l’Imprimerie nationale et bibliophile, demande en 1900
Yes, if you could scan the page(s) from the Imprimerie specimen book, that would be great. I'm working from the 16pt from the "Debuts de L'Imprimerie en France" book mentioned above (have no idea how I mistranscribed it as "Origines"). It has a fairly complete set of sorts, but is missing miniscule w and magiscule X (or, possibly, I overlooked their occurrences).
And thanks for pointing me to the IN specimen book; I wasn't aware of it. It's high on my want list now.
Also, if people are really curious to see my work-in-progress, ask and I'll post. I'm kinda excited about this one.
It's 600dpi, and cleaned up -carefully- to reduce filesize & get rid of the stuff on the back of the page. This is (almost) the best I could do - the repro in the book isn't so hot.
> have no idea how I mistranscribed it as “Origines”
Well, that is the title of a predecessor to the 1905 edition...
So you're doing a literal revival? Like you're not going to improve that "g"? It looks like a "y" that's a bit too happy to see its righthand glyph...
Thanks. I was hoping it was a more complete showing, but at having all the sizes is helpful. I see that what they have listed as a "14" is what I've measured as 16. Sigh, one of these days I'll actually learn to measure ancient sizes.
So you’re doing a literal revival? Like you’re not going to improve that “g”?
What I've drawn so far is quite literal, but I do plan on improving some of the glyphs. The cap "T", for example, reads too much like an "E". I'm not yet convinced I need to redo the "g", though. It's perfectly legible, and is a very distinctive feature of the font. It's also historically grounded, having the same shape as in Collard Mansion's 1484 Ovid Metamorphoses (arguably a more direct influence on this font than Gutenberg; in particular, the final letter variations are almost a direct match). [plate 22 in Morison and Day]
What would you suggest bastardizing it to?
> one of these days I’ll actually learn to measure ancient sizes.
1) I don't think there's a reliable way anyway. The talus (internal leading) can be anything, even negative, like if there's kerns at the bottom of the descenders of a script face - and it doesn't even have to be a script face: I remember seeing smaller sizes of some Baskerville with a "g" that extended below the Em! (Or maybe it was Palatino.)
2) Don't forget to account for Didot versus Postscript points.
> What would you suggest bastardizing it to?
I'd just make the top of the "g" more integral, not so slapped on. But only if it's not a literal revival... in which case I might not stop at the "g".
in which case I might not stop at the “g”
the beauty of [[OpenType]]: alternates.
Raph, I'm really curious about the letter "s". Would you change that at all or leave it as is?
This is a dodgy face -- it's a dog's breakfast with colour issues.
The way it is here, the designer probably figured he had a good balance of disparate elements ("goth remix", as Hrant put it), so normalizing problem characters kind of defeats the whole purpose of the font.
Raph, why not do your own remix from a selection of old fonts?
Nick: can I quote you in the marketing materials for the font? More seriously, I'm digitizing this because I think it makes an interesting statement, and given the resurgence in blackletter faces, I think it may well find niches in certain designs. I find it more readable than most blackletters, good enough for medium-length runs of text.
hrant (re size): yes, the talus of this font is especially confusing because the cap "V" has an ascender that vastly exceeds the normal ascent, so much so that it could almost collide with the previous line's descender. Indeed, in one of the larger sizes (I think 16/18 but haven't measured it carefully) you get the feeling that the compositor adjusted the line breaks to avoid just such a collision.
andigrl: actually, in the clip I posted, all the "s" glyphs happen to be final, so you're seeing only that form of teh letter. The other form is far more standard. I'll post three images of it here for your your amusement.
First, a raw scan (normalized for contrast) of one instance. This is a 2400 dpi scan, so it's about 467 ppem. The printing quality is fairly good for 100-year old letterpress, but of course you can see plenty of paper texture and a bit of uneven ink spread.
Second, I averaged together dozens of instances to form a composite image. I believe this image reveals very nearly the true shape of the metal. You can see a slight lightening of the interior of the letter on corners, indicating that some instances are chipped. This lightening always happens on fragile parts of letters, and the degree of lightening corresponds to the probability of damage. Similarly, the zone of ink spread is a light-gray blur emanating from the body of the letter.
Last, my digitization. You can see that I followed the original shape quite literally, even to the point of different rounding radii for the different corners. I may want to replace some of these rounded corners with sharp ones, but I also don't want to change the character of the font too much. In particular, at large sizes I find a mixture of soft and sharp corners to be unappealing.
>Nick: can I quote you in the marketing materials for the font?
Yes, for a small licensing fee.
>I averaged together dozens of instances to form a composite image. I believe this image reveals very nearly the true shape of the metal.
That's like averaging all the portrait shots in a photo shoot and saying that the composite best captures the subject. An interesting modern art idea, intellectually intriguing but dull in practice. There are always one or two shots that stand out, capturing the presence of the subject, far more attractive than the others, or a composite of them can ever be.
It's also like digitally remastering an old music recording by averaging transcriptions from vinyl. It would be better to work from the original masters, or snip the hiss and clicks. You can keep them as a featured artefact (Portishead, Founders Caslon), but averaging them just creates soft noise.
With letterpress typography neither the type ("the true[?] shape of the metal") nor an average of impressions reveals the nature of the endeavor: there is an optimum printing. The question is, what is it? That's open to critical interpretation (how much bite, how much ink, etc.) What was the type founder's intent? What did he consider the best impression? Was there something that he was aiming for with the typeface that he was never able to achieve in letterpress, that may be possible digitally? I suspect he would have opted for more sharpness, rather than less.
Repurposing a typeface by digitizing invlolves a critical attitude, a value judgement. The "averaging" method is one such judgement as to what is appropriate, but it is non-commital, a statement too fuzzy to be disproved. Not the same as the truth.
Ultimately there is no truth, only design.
> Yes, for a small licensing fee.
Yes, and don't forget to memorize the EULA verbatim.
> That’s like averaging all the portrait shots in a photo shoot
No, because a metal sort is not a person.
If the point is to make a literal revival, and when you don't have access to the original metal (plus access to a way to actually print with it in an "intended" way) Raph's method is supremely useful*, sort of because it removes printing variance. This, even (in fact especially) if the printing at hand was done in a way that the original designer didn't want (not that all of them were sophisticated enough to worry about that) because the "average" form is still something more than nothing that you can work from, to interpret further.
* It won't be long before it gets a lot more attention. Raph, send it to APHA for one.
Whether it's a good idea to average an image seems like a really pleasantly complicated question. On the one hand you could be seeking an average because you want to find out what it is - just curious - or because you think that it is meaningful on an artistic level in an of itself. Maybe valid... Maybe.
You also might want to use it as Hrant said, as the base to jump off from. That seems fairly viable to me. But my intuition is that if you are not careful a blurry image may lead to soft blurry thinking. Not that it has to be that way - but the averaging could be a way of letting a process be a substitute for looking & thinking. A dubious & weak substitute. It might well be a good tool - I can't say - I haven't tried it. But it does seem like it could be a kind of crutch you don't need getting in the way of the real job at hand.
I think that is probably why Nick was suggesting that a new remix might be a better route. Making your own would require a more direct & less automatic relationship with the shapes.
Being really litteral in your revival might be best served by picking an specific imprint to use & improve from ( plus some altrenates to think about ) gives you a more specific outline to develop reactions to - more specific aspects, more personality or character to like or dislike - and to make conscious decisions about.
And picking a specific instance ( or several one at a time ) also brings you face to face with the issue of ink/imprint/paper VS screen raster & laserprinting & other modern printing/output. There have been many Threads here on Typophile that talk about the relationship. The book ' Counterpunch : making type in the sixteenth century, designing typefaces now ' by Fred Smeijers has a great section in which talks about the way that laserprinters and other modern printing differ from metal print and what he regards as the mistakes that were made when classic faces were digitized. If you are making a revival it seems like this issue aught maybe; to be issue #1. Do you aready know about these issues? Am I preaching to the choir?
Hrant, What is APHA?
I hope your project goes well. This face has a ton of personality & verve. I think thats why I worry about averaging. It think that verve might be lost in the process.
They recently had a article by Kay Amert about computer-aided letterform analysis; but Raph's stuff puts that firmly in the shade.
In re-reading all this I have to admit I think that Nick is probably right about where the excitement is - in the value judgement.
Hrant - do you say this based on what is posted here or with additional material. I looked for your reference - This is as far as I could get without interlibrary loan's help -
"How many times have APHA members compared two typefaces, thought them similar (if not identical), and felt at a loss to prove it? Such a problem was faced by Kay Amert, Professor and Director of the Typography Laboratory at the University of Iowa, when she examined the types used by two French printers: Simon de Colines and Robert Estienne. She perceived an eerie likeness when before her lay two books: the former’s La Dissection and the latter’s Georgics-like poem La Coltivatione. A bit of a mystic, she cracked the problem in a dream in which the gestalt of a single letter floated before her computer screen; and, lo, superimposition became the answer. By bit-mapping a particular letter from one page in vertical lines and its equivalent from the other in horizontals, she saw consonance where they created a cross-hatch and differences where single lines extended beyond the cross-hatchings. When she presented still another comparison, some grew impatient to learn her views as to why there should be such differences: punch modifications? re-cuttings? variations in inking? make-ready? paper-dampening? But Professor Amert held that such questions called for other seminars and that her purpose was to recommend – to serious type-history scholars – the computer as an aid in addressing problems which have plagued them for centuries. - - Greer Allen"
That thing - no? What do you think make's ralph's analysis better? The part about the 'broken/weaker' bits?
Anyway - I have noticed that there is always a groupo of folk who hold that systems & analysis don't help the design of letterforms much or at all; and those that think they can be of service. Hrant, I think you Ralph & I belong in the second group as I see it. But I have to admit that even as I seek out systems ( Like DeGroots' letterform weight model) I remain slightly skeptical of them too and I think that's prudent even healthy - don't you?
This is an interesting bastard of a mix (the original, not Raph's methods…). Looking at the sample in the top post, I'm struck by three things.
1. That lowercase g has a top that is so [[Schwabacher]].
2. But that L is almost a [[Civilité]]
3. The type seems to have no real rhythm of shapes or internal elements, the way that most "blackletters" do…
The art nouveau era was symbolized by a number of attempt to reform blackletter by mixing it with itself or with foreign styles. To my knowledge, none of the attempts "worked"*
Raph's find doesn't either, but it could always revived it anyway. (I don't know what a user could use it for… the typeface didn't successfully capture any contextual spirit in its own time, either.)
Call me a [[designer]], but I think that the exact (or as close to exact as possible) recreation of someone else's intent is totally boring. Put those outlines where YOU think they should go. It isn't really important where the typefounder put them, I think. This face is not that "good" anyway. If you approach it more from a "designists" point of you, you might actually improve upon the past, rather than just recreate it.**
* After the second world war, some designers—mostly in Germany—had figured out a way to make blackletter work again, and it was not any of the art nouveau methods. Their method was The Pen. Of course many art nouveau forms were formed by the pen, too, but not with "consistency." Those 1920s faces were more legible, even if Hrant doesn't like pen-based forms.
**Not everything that existed in the past deserves to be revived, just because it happened to have been created once. In the end, you seem to be reviving this because you like it. But if people had "liked" this face, then it would already be with us in digital form. Show all of those nay-sayers what they were missing by finally making this design work! Its old forms will mostly likely make people think, "you know, maybe there was a reason that this had gone so long without having been revived…"
>No, because a metal sort is not a person.
That's why I said "like", and also mentioned another analogy, old recordings.
No analogy is exact, or else it wouldn't be an analogy.
The analogy is valid: there is a variance of instances in a photo portrait session, just as there is a variance in letterpress type impressions of the same character in a sort. The point is that in a session of "shots" or "instances" where the image is variable, there is always going to be an image capture that is truer than others, and hence also truer than an average. That's the way people see things.
It is not a scientific truth, but a subjective truth: the discriminating eye of an art director chooses the "selects" for a magazine, and the discriminating eye of the type designer should likewise make its decision. That is the responsibility of design. Averaging is a cop-out.
Having said that, I do find certain kinds of averaging fascinating, but in those cases there is some edge to the procedure, for instance the musical work of Brian Whitman, where he mines massive databases. The equivalent in typography would be a blackletter mix-font that averages every font keyworded "blackletter" at Myfonts.
> if people had “liked” this face, then it would already be with us in digital form.
I think that's too simple.
> there is a variance of instances in a photo portrait session
Because a human face is a dynamic, organic, living thing.
A metal sort however is a static thing, consicously designed. It's a lifeless lump.
> Averaging is a cop-out.
No, in this case averaging is a tool, a useful one, because it shows you what some other designer did, in a way otherwise inaccessible. Other designers had a brain too. But, again, this is only if you're doing a literal revival. Which makes me ask: why don't you go ahead and state a broader belief of yours: that revivalism is a cop-out?
> The equivalent in typography would be a blackletter mix-font
> that averages every font keyworded “blackletter” at Myfonts.
Fortunately though the 90s are over. Whew.
>A metal sort however is a static thing, consicously designed. It’s a lifeless lump.
But the process of letterpress printing introduces all kinds of unpredictable effects, due to things like varying wear of different pieces of type which represent the same letter, variance of inking, variance of pressure, and variance of paper stock. So how is one to know which is the "true" impression, any more than one is to know which photo captures the "true" personality of the sitter?
>No, in this case averaging is a tool, a useful one,
I agree, it can be useful, but it has to be a starting point, or a stepping stone, not the destination (or almost-destination)
>why don’t you go ahead and state a broader belief of yours: that revivalism is a cop-out?
That's not my belief. My belief is, and has always been, that revivals have their place in contemporary typography. I have done many revivals, and am working on one at the moment, actually.
What I have argued against is the suffocating dominance of revivals and old typefaces in our type culture.
>Fortunately though the 90s are over.
Generative art that uses algorithms to mine databases is a more recent idea, derived from Google, perhaps, which coincided with the millennium.
A similar "algorithmically driven" openness to unpredictability of form is present in LettError's work, e.g. Beowolf and the Twin typeface.
> But the process of letterpress printing introduces all kinds of unpredictable effects
Which is exactly what Raph's process nicely removes. That's its -obvious- value.
> any more than one is to know which photo captures the “true” personality of the sitter?
Obviously no photo can ever capture anybody's personality. But the shape of a piece of metal is on a qualitatively different different plane of definition. That's why the analogy is more confusing than insightful.
> it has to be a starting point, or a stepping stone, not the destination
If you're making a literal revival, it's more like a cornerstone.
> Generative art that uses algorithms to mine databases is a more recent idea
Yes, Art. Not Design.
> A similar “algorithmically driven” openness to unpredictability of form
"Anybody who intends to generate randomness by arithmetic means is, of course, in a state of sin."
- Johnny von Neumann
BTW, I can call him "Johnny" because his student was my teacher. :-)
Could I ask you how you averaged several different s's into one composite image? Did you do it in Photoshop, or is it some mathematical process that is way over my head?
I'm excited to see how the type will turn out!
Coming back to the revivalism angle:
> My belief is, and has always been
Ah yes, Perfection.
> I have done many revivals, and am working on one at the moment, actually.
> What I have argued against is the suffocating dominance of revivals and old
> typefaces in our type culture.
If you're talking about commissions, that's entirely different. Being paid to explicitly do something is partly excusable (if still somewhat of a cop-out; but we're all human). However, free-wheelingly contributing to the "suffocating dominance" is a major cop-out, no?
Also, I wonder, what do you think about type designers whose oeuvre is 90% revivals?
Maybe your thoughts on revivalism have indeed not changed since the Paleozoic; but if not, then your vitriol against revivalism has certainly subsided, just in the past half a decade or so! Wow. But why?
Antiphrasis: it's certainly possible to do it in Photoshop (I used Gimp myself), but tedious to the extreme in quantity. Thus, I hacked up some code to segment, classify, and blend automatically. The particular run that resulted in this "s" image was taken from a half dozen pages, contained about 10,000 total glyph impressions, and took several hours of computer time to crank through. If motivated, I could probably make the tools an order of magnitude or two faster, but even so it's a huge improvement over doing it all by hand.
Nick said, there is a variance in letterpress type impressions of the same character in a sort.. Yes, and that is simply not the case with standard digital type. (you could have a font with lots of alternates, perhaps even randomly selected, to simulate the flavor of letterpress, but that would be a whole other ball of wax).
Hrant and I had some email exchanges about Baskerville* specimens a few months back, and I pointed out an "a" impression that looked quite garaldish (garaldean?). If I were doing a Baskerville revival, it would be tempting to use this particular impression as a model. However, it would be a lot harder to argue that this form should be considered historically significant. I do think that the averaging technique, could help unlock some of the secrets hidden in these seminal sources.
*I had the pleasure of leafing through a Baskerville Aesop today, at Serendipity Books in Berkeley. Only $475 too!
I had said the true shape of the metal, but want to take that back. The shape of the metal is but one truth, and arguably not the one closest to the intent of the type designer or printer. Metal type is designed for an environment in which there is always gain on the press; the fact that better quality presswork often meant less gain does not mean that the designer's intent would have been satisfied by a perfectly sharp and gainless reproduction of the metal shapes.
What I would like to claim, instead, is that this averaging procedure produces an image that is very nearly as clear as a high-res capture of the shape of the metal. A straight-up average is going to be a bit blurred, largely because it's impossible to totally get rid of misregistration errors. But, that said, the very low noise of the result image lends itself well to sharpening (or, for an extremely sophisticated practitioner, deconvolution). I'm playing with this now, and the initial results are fascinating indeed. You can actually see the body of the metal being surrounded by the ink gain.
So yes, I would stand behind the claim that this averaging technique is an analytical tool of unsurpassed power. Further, that these images are arguably better than either going to the metal or choosing individual impressions if one's goal is creating a faithful reproduction. Whether there's any artistic or design merit in doing so is another question altogether, and I'm heartened to see the debate over it.
That sounds really interesting... it reminds me of a program I created in my neural networks class a few years back. It analyzed (b/w) pictures of human faces and was able to isolate the eyes and save them separately, and with more training the program got more and more efficient. I would assume that good OCR programs use neural nets to analyze glyphs... but there might be other ways to do it as well. Thanks for the info!
>then your vitriol against revivalism has certainly subsided,
Think globally, act locally -- that means local culture providing meaningful, profitable work for people everywhere, work which expresses the diversity of culture, engaged with today's issues. Typographic culture falls far short of that: the typefaces most people in the West use were designed long ago and/or far away, provided by trans-national corporations. There is a role for continuity with the past, but it's overdone.
I hope that's not vitriolic.
>what do you think about type designers whose oeuvre is 90% revivals?
I'm not into finger-pointing at individuals, only at TNCs, which are like public institutions.
I've argued, on other Typophile threads, that revival is not type design, because it's taking credit for someone else's work. But it's not a black and white issue, there are many shades of originality.
I don't think I've eased up on my cause, after all I'm onto Raph's case here. But I have nothing against the guy, it's just a philosophical discussion. He made certain statements about "truth" which I contested, hence the debate -- no vitriol.
>further, that these images are arguably better than either going to the metal or choosing individual impressions if one’s goal is creating a faithful reproduction. Whether there’s any artistic or design merit in doing so is another question altogether,
This approach: scientific analysis accurate, art marginalized as a gloss.
It's not possible to separate the art and technology so neatly. The question of what is faithful is a question of design belief.
For instance, there is a quality of sharpness which would be apparent in an individual impression, which the average is not faithful to.
The scientific assumption is that what can be accomplished with a bit of technology is better than what can't.
>"Anybody who intends to generate randomness by arithmetic means is, of course, in a state of sin.”
I said unpredictable, not random. A database that constantly changes from day to day is unpredictable, because one cannot predict the future. The appearance of the Twin font, driven by the weather in Minneapolis, is something that von Neumann would have approved as being truly random.
> You can actually see the body of the metal being surrounded by the ink gain.
This is really interesting to me - could you elaborate?
> revival is not type design
> The question of what is faithful is a question of design belief.
This approach: art (expressing myself) is what counts, scientific analysis (or just plain thought, really) an impediment. That's not Design. Fear of analysis, minimization of technique, these things prevent progress. In a way, shunning analysis is a form of revival, that of ideas, approaches! That is most certainly not type design. This field is littered with people who really don't have the intellectual stamina and curiosity to exceed the styles -but also ideas- of the past.
> there is a quality of sharpness which would be apparent in an
> individual impression, which the average is not faithful to.
I don't think you understand what Raph is doing.
Averaging isn't necessarily blurry.
Are you perhaps simply seeing in this the things you need to reject it/analysis?
> The appearance of the Twin font
I was talking about Beowulf.
>This approach: art (expressing myself)...
That's not what I had in mind. I'm talking about what I do, which is commercial type design, where the art and design involved are hand-eye-mind skills. This is what I'm opposing to a revival that leans heavily on automated techniques.
I was thinking more of Petr von Blokland's statement "What can be automated will be automated, what cannot be automated is design." That defines the disputed territory. I don't like to see subtle design skills replaced by crude technology. It's all very well to tout averaging as a great new tool, but the averaging idea is not new. I believe Adobe Garamond for one, uses it.
Automated averaging, used as a design platform, is a blunt and facile instrument. As an analytical tool, it really doesn't help the (re)designer understand the typeface in the way that looking at a variety of instances and drawing one's own composite does.
>Averaging isn’t necessarily blurry.
Well, I said it wasn't sharp. Usually sharp is contrasted with blurry, but in the case of averaging type outlines, what one ends up with is not a blur, but a hard edged blob. It's not blurry, but it's not sharp, because it has soft corners.
>I don’t think you understand what Raph is doing.
>Are you perhaps simply seeing in this the things you need to reject it/analysis?
Please, lay off the ad hominen comments.
I'm out of this thread.
> This is what I’m opposing to a revival that leans heavily on automated techniques.
But that's because of your opposition to literal revivals.
You're confusing the separation between the intent and the methods.
> "... what cannot be automated is design.”
That glosses over the difference between Art and Design that I'm trying to address; the difference between self-expression and serving others. The former hates analysis, the latter requires it. Considering everything else I've read you write, I suspect that's the crux of your rejection of Raph's effort. All I'm trying to do is understand, everything, yes using personal contexts too. And I do this to improve myself... I need that badly! I'd like to understand why people hate/love the things they do, and make the fonts they do, sort of like reverse-engineering*, and grasping what's in their heads is a crucial key. And I think there's a strong parallel here with revivalism: if you get an idea of what the old dead guy was thinking (which is most apparent in the sorts, not the sloppy printing of them) you will do a much better job.
* "Yucky engineers! All they do is ruin our precious painti... I mean typefaces!" ;-)
> I don’t like to see subtle design skills replaced by crude technology.
But this isn't about that. Anybody making a revival should be interested in the original. And this technology helps in that. It's really quite obvious. It doesn't replace, it guides, that's all. Think of it as an astrolabe.
> the averaging idea is not new.
Well of course it isn't. Nobody is saying that. Raph's work however is a quantum leap.
> it really doesn’t help the (re)designer understand the typeface
And what does, shrooms?
> It’s not blurry, but it’s not sharp, because it has soft corners.
Repeat: I don't think you're undestanding the technology, and the way to best apply it.
> lay off the ad hominen comments.
It was a question! Just answer it (if you like). Please explain to me what I don't understand about your stance. Assuming you are interested in defending your views (where a tiny bit of self-doubt would help too), as opposed to merely stating them.
I was afraid you'd pull a whine-and-split... How is my questioning more insulting than "revival is not type design"? You have shades? I'm Mister Shades! :-) As much as I don't have much of a taste for revivalism (and it's quite sad that you're trying your darnest to make sure not to appear to be agreeing with me on that), I embrace its utility: it puts bread on the table (repeating the past/present always does) and it's a form of celebration (and society needs that); but just as important is theory/progress, and this is an area where controlling one's fear of technology is crucial.
>“Yucky engineers! All they do is ruin our precious painti… I mean typefaces!” ;-)
Is this designist comment really necessary?
Why should I continue with this thread if you're going to fill it with macho "humour"?
>How is my questioning more insulting than “revival is not type design”?
It wasn't questioning.
In the first place, you said I don't understand what Raph is doing.
Then you suggested I'm seeing things in it that aren't there.
I'm not prepared to continue the discussion after being forced into that kind of corner.
"Revival is not type design" is not a personal comment, it's a statement of fact. For instance, Hnaffe is the designer of Christian Gothic, not Raph.
> Why should I continue with this thread if you’re going to fill it with macho “humour”?
Macho? I'm not macho. Unless you think you're a pansy. I don't think you are. And I would think you prefer humor to insults. I'm trying to work with you here; but I'm not a cyberpsychotherapist; I could never answer that question for you.
> you said I don’t understand what Raph is doing.
A certain aspect of it, yes. You clearly understand the gist of it, and it seems that this is enough to prevent you from trying to understand the details, due to what seems to be an aversion to Analysis?
> “Revival is not type design” is not a personal comment, it’s a statement of fact.
Huh, and I thought there were shades.
> However, it would be a lot harder to argue that this
> form should be considered historically significant.
As Peter hints, revivalism is complex.
I would ask: Who is the font for? Other type designers? Quite unlikely. What is for example THE Baskerville for the contemporary user? Is it JB's original? No way - virtually nobody has seen that. I would argue that it is more the impression that people have accumulated over time (a lot like what Nick wrote in another thread) from all the Baskervilles they've seen. So for example take the "a" in Monotype's Baskerville. I find it quite ugly; I would never make that on my own. And John Hudson has argued (on the ATypI list) that it's not necessarily the historic form. But really, it doesn't matter: it's what people know, and my commission is to provide something archetypical. So my revival will have that ugliness (just not exactly the same). To me, that's good design.
>You clearly understand the gist of it, and it seems that this is enough to prevent you from trying to understand to details, due to what seems to be an aversion to Analysis?
Hrant, this is the crux of the ad hominen idea that was in the other thread recently.
You shouldn't pursue an argument by saying that the other fellow doesn't understand the topic and you do.
>Macho? I’m not macho.
I didn't say you were macho. I said it was a macho comment.
As are your frequent quips about "artistes" expressing themselves.
>Unless you think you’re a pansy. I don’t think you are.
That's not my idea of a compliment.
“Revival is not type design” is not a personal comment, it’s a statement of fact.
Nick, I beg to differ. It's simply your opinion.
First, revivalists are never given a complete Adobe character set. Foundry types almost universally never include daggers; section marks; braces; commercial ats; math operators; currency symbols other than those of the country in which the bodies were cast; or oe and ae dipthongs, eths, thorns or German double-esses. In order to include those characters in a font, they have to be designed.
Second, most revivalists credit their sources, when they are known: no one is trying to hoodwink anyone here. As I have stated before, the whole idea behind revivalism is to make typographic resources available to a computerized world which heretofore existed only as metal type or handlettering on a page. You might not consider the process to be creative, but it is transformational, and it requires patience, judgment and an eye for design.
> You shouldn’t pursue an argument by saying that the
> other fellow doesn’t understand the topic and you do.
Pursue an argument? I'm communicating my view that you're missing something - and I just said you understand the gist of it just fine - the details however are important. Are you listening to the communications, or just blasting a deeper trench?
> I didn’t say you were macho. I said it was a macho comment.
But I said it. It came from my head. There you go again compartmentalizing things, removing context. Although you didn't do me that favor when I once wrote "that makes you sound stupid". (You see, Nick, it won't work - not with me.)
> As are your frequent quips about “artistes” expressing themselves.
Well, I think they're a problem. Would you rather I dispense with the humor? I doubt it. Would you rather I shut up? Well, of course. But that ain't gonna happen.
> That’s not my idea of a compliment.
Fortunately, I'm not out to compliment you.
Some of your fonts sometimes though, yes.
> It’s simply your opinion.
No, it's what the word "design" means.
There is a distinction between someone coming up with the idea for a typeface, and realizing it in some form or other, and someone else building on that idea.
The originator designs the basic pattern of a typeface, and subsequent adapters and revivalists follow that. Following a design is not designing, it's interpreting. Nothing wrong with that, but it's not design.
Bob Dylan never wrote any guitar solos for "All Along the Watchtower", but that doesn't make Jimi Hendrix the composer.
> The originator designs the basic pattern of a typeface
You speak of a mythical beast, which exists only in some people's imaginations.
It should be obvious that such "origination" is merely an illusion.
>It should be obvious that such “origination” is merely an illusion.
Either you see it, or you don't.
No, it’s what the word “design” means.
Nick, I would humbly suggest that the word "design" has a certain meaning to you; unfortunately, your meaning is not the only meaning the word can have . I would also suggest that you check out the Oxford English Dictionary, the ultimate reference source for the English language, to see exactly how many other possible meanings the word has, very few of which agree with you. If you cannot, for a moment, accept the fact that there may be other valid definitions of the creative process, I would further suggest that you lay off the nose candy, or whatever else it is that gives you the idea that your words carry the same authority as, say, the Ten Commandments.
Concise Oxford English Dictionary
1 a. a preliminary plan or sketch for the making or production of a building, machine, garment, etc.
b. a scheme of lines or shapes forming a pattern or decoration.
c. a plan, purpose, or intention.
4a The general arrangement or layout of a product.
b. an established version of a product."
Nick Shinn's distinction between a design and various interpretations does seem to have a basis in the dictionary meaning. Of course, the distinction between a design and an interpretation or realization of it is not a sharp one.
I think Christian gothic follows a singular, complex (though definable in simple terms) basic pattern, and it can be effectively discerned and described with the help of Gerrit Noordzij's type-descriptive protocols. An understanding--and faithful digitization--of Christian gothic might benefit as much--if not more--from viewing the letterforms through Noordzijian glasses as it will form mechanical averaging. The basic pattern might be less shakely a product of the historically retro(pro)jective imagination than a putative set of Fleishmannian ideas on readability.
>I would further suggest that you lay off the nose candy, or whatever else it is that gives you the idea that your words carry the same authority as, say, the Ten Commandments.
Oldnick, I thought you rejected that sort of personal attack.
William, please maintain focus on the content. (Was that Western enough?)
Nick, I'm sorry the topic of Is-Revival-Design came up on this thread (Hrant's to blame), because Raph has quite rightly said he is digitizing the face, not designing it.
>The basic pattern might be less shakely a product of the historically retro(pro)jective imagination than a putative set of Fleishmannian ideas on readability.
That makes a lot of sense. But it is still a face with spotty color. For instance, "e-n" and "e-r" don't have the same space between them. Maybe what it needs in the digital domain is either some fancy kerning, or alternate forms (courtesy of OpenType). I can't accept that the poor color is a Fleischmannian necessity.
Yes, it's all my fault. Even the part about you trying to hide your disdain for the oeuvre of many accomplished type designers, including some you've recently colluded with in attacking me. It's all my fault! Grab your pitchforks and torches!!
> But it is still a face with spotty color. For instance,
> “e-n” and “e-r” don’t have the same space between them.
Does anybody literally-revive spacing, at the micro level?! Besides Justin Howes I mean. Maybe they should? But you certainly can't think so.
> fancy kerning
Why would it need more than good basic spacing and simple kerning?
> I can’t accept that the poor color is a Fleischmannian necessity.
Wait, I don't get it. Are you saying that Fleischmann's work has poor color? At the font level?
And what does that have to do with this again?
BTW, I'm glad you changed your mind about leaving.
William, my words were very carefully chosen and they were not an attack on Nick, only his definition of "creativity." I am not dissing the man, just challenging some of his mistaken notions, which is what lively discussions are supposed to be about.
Unless one possesses the power to say "Let there be light" and light does indeed come into being where it did not exist before, nothing a human being does is creative. Rather, if a mere mortal "creates" something new and unique, it is merely an act of synthesis, building on what came before. And nowhere is this fact more true than in the area of typeface design. The basic rules of letterforms are fixed: the design of an A, no matter how unique, must suggest, unequivocally, an archetypal A to the viewer of same, or the design is utter worthless.
Equally, unless one is prepared to state--also unequivocally--that any nuances one may devise for an alphabet have never, in the entire history of mankind, ever been conceived of before (which statement must, by necessity, proceed from Divine omniscience), then one cannot legitimately claim to be a "creator," when lesser mortals (here read, mere revivalists) are not.
Finally, Nick: enough of the Hrant bashing. I believe that Hrant would be the first to admit that, at times in the past, he has gotten a little edgy and insulting, but...I believe that, within this particular forum, he has been restrained and has offered a lot of constructive criticism and helpful insight. You don't have to cut the guy any slack: just give him his due.
>Equally, unless one is prepared to state—also unequivocally—that any nuances one may devise for an alphabet have never, in the entire history of mankind, ever been conceived of before
Now YOU're sounding Hestonian, Nick.
It's not about nuances. It's about resolving a unique set of design principles.
But as a matter of fact, I have devised quite a few never-before conceived nuances.
Not that they necessarily have any great typographic value, but they are original.
>trying to hide your disdain for the oeuvre of many accomplished type designers,
Sorry, is my disdain showing? But really, I don't understand why you'd think I have disdain for people who do revivals. There's plenty to admire. Sure, I respect original typefaces more, but everyone has their preferences, that doesn't mean they detest what they don't adore. Also, while I occasionally deplore the general state of affairs (the glass half empty part of it), I'm not going to point fingers at anyone (except of course font bundlers), least of all indie foundry designers. When I say "revival is not design", my purpose is to respect the original designer of a typeface -- not to insult the revivalist, but to question their (or a third party's) use of the term "designer" for the bezier-wrangling.
>Does anybody literally-revive spacing,
Look at the pairs I mentioned: the extra serif on the foot of the "r" means that "e_n" will create more of a hole in text than "e_r". This could be solved by kerning, or by a contextual alternate form of e with a longer tail.
>Are you saying that Fleischmann’s work has poor color?
No, but if Christian Gothic has a divergence of letterforms, for Fleischmannian reasons, that's problematic (as is often the case when diverging from the norms of centiuries) for text colour.
> I have devised quite a few never-before conceived nuances.
1) You don't know that.
2) Nuances are such a small part of type design, at least in terms of progress. Think more about the structures. But to do that you need... analysis! Not mere expression.
3) Nuances can exist just in the "simple" act of revival too.
> is my disdain showing?
Not any more. It used to. You seem a lot more careful now. Maybe it's part of becoming mainstream?
> This could be solved by kerning
Well sure (if the whole is tight enough). I was wondering about your "fancy"... How many pairs do you typically define for let's say basic Latin?
> if Christian Gothic has a divergence of
> letterforms, for Fleischmannian reasons
What are these reasons, exactly? I hope you're not sticking with the view that he was an Emigre-style designer. That would be a reason for his descendants to get pissed. :-)
> that’s problematic (as is often the case when diverging
> from the norms of centiuries) for text colour.
No, structural divergence (the important, interesting stuff) does not preclude even color.
I'm going to gently disagree with Nick on a number of points.
First, I believe my averaging technique does produce sharp results, both in the sense of edge acuity and curvature at corners. Keep in mind we're looking at a 14pt glyph here; on a typical 96dpi screen, that means we're looking at 25x magnification, or 50x for the last image I posted.
Second, your e_r vs. e_n argument doesn't convince me that there's a color problem, for two reasons. For one, note that the left serif of the 'r' is shorter than the left serif of the 'n', so the stroke is closer in (the distance between the right tip of the 'e' and the left serif is pretty much the same). That, in turn, balances the perception of greater space due to the lack of a projecting lower left serif on the 'n'.
Also, ideal spacing for a printed page may well not be ideal spacing for an enlargement viewed on screen. I've proposed before that as size increases, gaps become relatively more important, and distances between stems relatively less. So even if e_n still looks a bit gappy to you, that perception may go away completely at actual printed size.
There are significant differences between the various sizes of this font, and these differences have a definite impact on color. I very deliberately chose the 14pt size, mostly because I liked its color best. One of the most important tasks facing a "digitizing interpreter" (or whatever term seems most appropriate) is to choose between the various sources available. There's an unhealthy tendency, I think, to seek out large sizes for their sharpness and robustness to printing variations. The averaging technique has the potential to change this balance, and more clearly reveal the cool stuff going on in the smaller sizes.
You should be to see this fairly clearly in the last image I posted. In particular, at about 5 o'clock, at the thinnest part of the stroke, you should be able to see a dark-gray stripe nestled inside a blacker outline. I believe the dark-gray is actually a good image of the metal, while the outline is composed of ink spread. If so, hopefully I can unlock one of the last mysteries in my quest to understand ATF's optical scaling, to wit the relative contribution to thickening from the pantograph and the ink spread on printing. I'm looking forward to getting some time to refine my techniques so I can capture even clearer images - the ones I've posted here are still a bit murkier than I'd like, especially for visualizing these more subtle aspects.
In any case, hopefully we can dispense with much of the flamewar on this thread, as I am indeed not making any particular claims to creativity. The main reasons I'm doing this font are to learn more, and to refine my tools. I feel good about the progress on those goals even if this font is never used for anything other than the nameplate on the front door of my new apartment.
>> I have devised quite a few never-before conceived nuances.
>1) You don’t know that.
Yes I do. Admittedly I haven't seen every font ever made, but I feel reasonably sure that I would have come across certain things if they had been done previously. Such as an oldstyle italic monowidth unicase font, for instance. (A structural concept, but one requiring considerable nuancing to execute.)
>You seem a lot more careful now.
Hopefully one learns from one's mistakes. Type designers can be extremely sensitive to even the most objectively couched analysis, which coming from a peer can seem a bit high and mighty.
>I was wondering about your “fancy”
Just the idea that an e_n kern might be useful.
>structural divergence (the important, interesting stuff) does not preclude even color.
Well, that's the big challenge, isn't it.
>sharp results, both in the sense of edge acuity and curvature at corners.
I don't see that. In the original scan there are a number of sharp features, even at the low resolution. It may be uncertain as to which are parts of the design and which are artefacts (and averaging would differentiate), but nonetheless, there is a quality of sporadic sharpness (aka sparkle?) to the individual impression, which disappears in the average.
Also, consider that the quality of different pieces of type will vary, both in mint condition and through varying wear. An average is less than top quality. In the same way, certain imprssions will be "sweeter" than others, due to slight variations in paper and inking. Again, the average dulls the best quality, and it is up to the discriminating eye to discern what is best.
The e_r vs. e_n space: the absence of a lower left serif on the 'n" creates a baseline opening between the letters, so although the fit is slightly tighter, it allows contigual white space to interupt the x-height's evenness.
>So even if e_n still looks a bit gappy to you, that perception may go away completely at actual printed size.
I've designed many text faces, and took that into account.
>my quest to understand ATF’s optical scaling
Why don't you work with a face where you also have access to the metal type? Then you could compare that with the impression, rather than deduce it.
>hopefully we can dispense with much of the flamewar
I apologize for inadverently flaming you, as a result of my ongoing headbutting with Hrant. What you're doing is fascinating, and creative, and throws a number of philosophical design issues into the spotlight.