The "Ikern" enigma

Curioustype's picture

For those who are unaware, Igino Marini has developed what he calls "Ikern," a program/service/concept that generates kerning for type. According to most everything I've read about it, the entity that is "Ikern" works quite well. But it raises questions for me.

First (and should he visit this site perhaps he could provide a response), Mr. Marini has been particularly vague when it comes to, well, pretty much everything other than generic explanations of the hows and whys of IKern, choosing instead to focus more on the positive reviews given by designers for whom IKern has been successful.

Since these are designers I would tend to believe, I can only assume Mr. Marini has developed what to some could be an extremely valuable program/service/concept. The only problem is, as of today IKern is still very much the man behind the curtain. And as far as I can tell, the only real way to find out anything about it is to contact Mr. Marini directly via email.

Am I the only person who feels uncomfortable about this? All I really want to know are the essentials: is this or will this become "software," available for purchase at some point? If it's a "service" only, what are the costs? What is the turnaround time? How in general terms does the program/service work?

I also am one who believes Mr. Marini has created something that stands a chance of becoming quite successful in a variety of ways, but for my own personal reasons I simply don't feel comfortable going on fishing expeditions, regardless of how friendly the fish might be. Not just in this case, but at all. Therefore, does anyone have the kind of basic information on IKern I've asked above?

Additionally, if IKern was able to produce quality results - and I have no reason to doubt it does - would it be something you'd use? Would you buy it if it was released as software? I just find the whole thing intriguing but strangely mysterious.

Curioustype's picture

I've seen the site, and been through it top to bottom. That's kinda the point here - none of the questions I asked in my original post are answered on the site.

Ricardo Cordoba's picture

I was simply adding a helpful link, so that others might be able to get there and possibly comment on your questions. :-)

beejay's picture

step 1. contact Igino

I've contacted him about iKern, and he answered all my questions promptly via email.

Because he has created something proprietary, and because it's probably something that others might want to duplicate, I can respect his decision not to reveal all the machinations behind the curtain.

nothing mysterious or nefarious to report. :)

Curioustype's picture

I have an idea why he would be so guarded against disclosing some things, for which I do not blame him at all. However, some of the most basic information wouldn't be letting the cat too far out of the bag. Such as whether or not it's something he does himself or relies on software to complete, the costs, turnaround time, etc. I'm sure eventually I'll just email him but before doing that it would be nice to know at least some general information; particularly the cost of this service. I'd be wary of paying for this service now only to have his creation released as software and available for purchase at a rate less than the cost of the service. The secrecy of the whole thing just struck me as being strange, that's all. I can only wonder how many others are like me, who are not prone to inquiring about any kind of product or service without having more information.

beejay's picture

well, i'm a curious type, too. But a point to consider: Take a look at other people who do backend work on typefaces. How many of them post their rates on their site? I can't think of any. Would they want their rates revealed on a public forum such as Typophile? Would you want your rates and/or proprietary methods revealed here?

It seems respectful to keep that type of specific information out of this thread, unless it's provided by Inigo.

His site reveals quite a bit.

Further curiosity can be satiated with an email—no fishing expedition required. :)

k.l.'s picture

Re program/service/concept
and does he himself or relies on software to complete:
For the designer, it is service, not a program to be had. This is said on the website. Google a bit ("ikern"). Igino Marini has contributed to two or three Typophile discussions and answered this and other questions already.

Re generic explanations of the hows and whys:
As far as can I see, iKern is based on abstract ideas about spacing. A description of this tends to be "abstract" too. Yet the description is not abstract-as-business-talk but abstract-as-technical. (Designers hate abstraction, I know.) Perhaps it helps to know that English is not Igino Marini's first language -- nor is it mine -- and I know very well how hard it can be to explain conceptions that are far from trivial.

Re positive reviews given by designers
or, which seems to be your question, quality of results:
Different designers have different ideas about what good spacing is. Speaking for myself only, I can say that I was surprised how good iKern's results are. (To put it this way: After trying test words on the MyFonts site I think that iKern may help you improve the spacing of Morning Sans.)

Other issues have been dealt with by beejay and others already.

If you feel more comfortable with an application that you can install on your own computer, there also is KernMaster. Whether service or application or manual work, or a combination of two or all three, really depends on you, your expectations, or how much you want to control everything. The latter causes me to do everything by myself but I am aware that this is irrational and not always backed up by facts, and am trying to get rid of this attitude.

blank's picture

I never knew that KernMaster actually auto-kerns. And even if I have to use dollars to buy it, if the output is good, it’s worth the money! As I DTL product I expect it to be good, but would anyone care to comment about the quality of the output?

dan_reynolds's picture

It is sad to see so many of my fellow young designers come here looking for ways to automate or sub out fitting and kerning. These are just as essential to typeface design as drawing outlines. The way to properly space a font is to do it yourself. The way to learn how to properly space a font is to keep designing and spacing new fonts. Your first fonts are not going to be fit well. But whose first fonts have good outlines or lettershapes either? You will get better. Do not invest money trying to find short-cuts. Work hard. For a long time. It will be worth it in the long run.

dezcom's picture

Drawing the outlines the first time is not the last time. The interaction of spacing and kerning causes us (well, at least me) to revisit the outlines to delve into fitting and redrawing. In the process, I can see problems in the drawing. I don't know how to pluck one part of the process out and make sense of the rest. Perhaps, like Karsten, I want to control it all. I am still learning what makes the whole thing tick. I may never have the whole answer but I have always needed to know how things worked.


Mark Simonson's picture

That's a good point, Chris. I can't design a typeface without taking its spacing into account, and often adjust the shape of letters to address spacing problems.

I don't imagine Ikern, or anything similar, would be much help at all with a connecting script, for example. The only way to fix the spacing in a connecting script is to change the shapes of the letters.

Maybe it's possible to draw all the letters first, then space the letters, the end. To me, it's a loop, not a one-way street.

William Berkson's picture

Like Chris and Mark, I have found that modifying letters in the process of kerning is something I feel a need to do. This has been particularly true with the swash letters that are part of the font I am finishing. IKern sounds very good, but I think having a application such as FontMaster, that I can deal with interactively, altering letters if I want, is something I would be more comfortable with in the future. (This first time I felt I had to do it all to develop my eye; in future I would judge a computer-generated result better.)

Because I suspect a lot of designers feel like Karsten, Chris and Mark, I have a feeling that IKern might prosper better as an application rather than a service. But of course I don't know.

Curioustype's picture

I love all the interesting comments here. Among them:

"For the designer, it is service, not a program to be had. This is said on the website."

I realize it's a service in regards to interaction between Mr. Marini and a designer. My question was more along the lines of, does he go into a font and kern it after applying some script/formula/whatever, or has he created actual software for it? I'm thinking in terms of compatibility here.

"After trying test words on the MyFonts site I think that iKern may help you improve the spacing of Morning Sans."

No kidding ... I totally suck at kerning and spacing. But not for lack of lengthy, frustrating effort. What especially drives me nuts is the difference in the appearance of kerning just between two different applications. If I test the kerning in Quark, it's almost guaranteed to look different how it appears in Microsoft Works. And that's just kerning. Even checking the spacing in Notepad vs. Wordpad is mind-boggling. All of which of course fails to even mention how it all looks in print. In many ways it's like a house of cards. As an aside, I think you'll see an ongoing correlation between my confidence and knowledge in kerning with the cost of each released typeface.

"I can’t design a typeface without taking its spacing into account, and often adjust the shape of letters to address spacing problems."

I'd love to hear some examples of this. I would assume this also means you often adjust the spacing to address problems with the shapes of letters, correct? Just curious, because relying on one without the other, in my mind, makes one more important than the other.

"Do not invest money trying to find short-cuts. Work hard. For a long time. It will be worth it in the long run."

I work very hard and am definitely improving at it, if not due to sheer repetition. However I'm not sure I'd call effective automated kerning a "short-cut." I wouldn't call it the end all, be all either. At this point I'm still trying to feel my way around because I see so many different avenues both taken and available. For example, I've seen highly talented designers turn over an entire typeface to be kerned by another individual, who then does it manually and not by applying automation. Is that considered a short-cut? If I had a partner who was a kerning whiz but didn't like drawing letters, would teaming with him/her be a short-cut? There are all kinds of considerations and approaches; and while I continue to grind away at learning the concept as you suggest, ultimately my goal is discovering what works the very BEST, all things taken into account.

begsini's picture

wait, designers "hate abstraction"???

i'm a designer, and i love abstraction. value it. could not design without it.

Mark Simonson's picture

I’d love to hear some examples of this. I would assume this also means you often adjust the spacing to address problems with the shapes of letters, correct? Just curious, because relying on one without the other, in my mind, makes one more important than the other.

An example of adjusting the shape of a letter might be changing the position or size of the ear or the bottom loop on a two-story lowercase "g". When it stands alone, it might look great, but it might not fit well next to other letters, and cause gaps or crowding.

Kerning is meant to address problems with spacing letters with irregular profiles. Some letter shapes have unavoidably irregular profiles (T, L, Y, A, etc.). Sometimes these shapes can be altered to lessen the need for kerning.

I don't see how anyone can design a typeface by only considering the shapes of the letters in isolation. Characters in typefaces don't exist to be seen in isolation, they are meant to look good next to each other. Spacing is not more important than character design, or the other way around. They should ideally be considered simultaneously.

blank's picture

I don’t see how anyone can design a typeface by only considering the shapes of the letters in isolation.

I don’t see how using an automatic kerning tool implies drawing the letters in isolation. There’s no reason to assume that no preliminary spacing is being done by people working this way. I see no suggestion on the part of Igino or curious designers that automated process are not being performed in a vacuum and that if a pair doesn’t work when the spacing is generated, that the flaw goes unfixed. Just like when using a computer to interpolate weights, autospacing is still going to need watching. But it does allow for designers who can’t be font designers full-time to finish typefaces in much less time, and perhaps do a better job than they would if they can only fit in time between work/kids/etc..

Mark Simonson's picture

Okay, so that was a bit of a straw man.

When I started making digital fonts, I would draw the characters in Illustrator and then import them into Fontographer to do the spacing. I didn't realize until I started using FontLab and had to abandon working in Illustrator (because Illustrator import was seriously deficient back then--it's better now), that I was making a lot of work for myself by separating the drawing from the spacing. When I draw a font now, I keep the metrics window open and space the font at the same time I'm drawing it. When I'm done with the drawing, I've also done most of the spacing. The only spacing step left is the kerning. The only reason the kerning is done separately is because it has to be done systematically. If you do it as you go along, it's easy to overlook combinations. Plus, you can't know what the problem combinations are until you get the normal spacing and drawing finalized.

To me, it's all one organic process. When I'm finished, everything is just the way I intended. I love finding ways to save time and effort, and I use every short cut I can, but not on the design part. I see spacing primarily as a design activity, not a routine task I can leave to an algorithm.

I'm not saying there's no value to Ikern or other auto-spacing/auto-kerning services or applications. For some situations (maybe yours), it probably would save time and effort. I just can't imagine working that way.

iginomarini's picture

“The 'iKern' enigma” is a very good title. I will surely use it on my site: it will be a new section for the people who think it'll be worth reading it! I really think there is an enigma: why Mr.Miller has spent more time on this thread than he would have spent writing me directly? For the other readers I could copy and paste at this point: on the very first page of my site I wrote: “I’d be glad to let you test iKern”. After everyone's own eye, testimonies are what more nearer to the direct experience I can conceive. I would like all the type designers I worked with would describe their experience. Some did and I thank them. Some don't want to go public this way. And I can't blame them: there's no lack of judges. And people ready to put in a bad light other people's work.
Has anybody experienced the whole iKern's process so to say that feedback between metrics and shapes is impossible or adversed? Karsten tested iKern and his opinion is important for me. Who said that using iKern means designing without sidebearings in mind? In my site I wrote: the designer remains responsible for “having primed” the metrics themselves. iKern is just a tool: it won't add originality, style, expertise or mastery. It will serve you with the best metrics possible whenever you want. For your utility and advantage only. Usually I send 1 or 2 intermediate versions. Lately I sent 9 final versions in a period of 2 weeks. It's not to push a button.

To Mr. Simonson: obviously when letters have to touch, design and spacing are the same thing and it's not the iKern's game, but... One year ago I've been challeged to space a font with letters some touching and some not. In that circumstance I added instructions to selectively force contacts and change sidebearings accordingly so to have no kerning. That development stopped there because it was just a trick really out of the core concept and less elegant than just setting manually. About adjusting shapes for letterfitting purposes. I have in mind the Fell Types: a good result can be achieved even with some of the most technically wrong typefaces ever. Are you sure that, if one would dare to “optimize” them, they would still shine? You also wrote: “I see spacing primarily as a design activity, not a routine task I can leave to an algorithm”. I wrote on my site explanations, and I will add more and more, trying to make understand that iKern is like a car, not an automatic pilot. It's alternative to the feet, not the brain.

To Mr. Reynolds: You wrote: “The way to properly space a font is to do it yourself”. Do you think that the type designers fully satisfied of my work are less competent than you? How would you consider a type company willing to “invest money” in a “shortcut sad to look for” like iKern?

Igino Marini

Mark Simonson's picture

Fair enough. I didn't mean to criticize your software/service directly (which I admit I don't know much about). And thank you for clarifying what it does with your car analogy. I think I understand now.

I was mainly trying to say that you can't not think about spacing when designing a typeface. It seems foolish not to.

James Arboghast's picture

Spacing is part of typeface design and integral to letter design. That's how I work. I have no views on Ikern.

@igniomarini: Do you think that the type designers fully satisfied of my work are less competent than you?

What gives you that idea? You're making an ad hoc assumption there about the way Dan Reynolds thinks, and we would all be better off if you did not try to read between the lines like that or put thoughts into people's brains.

j a m e s

k.l.'s picture

In brief, Dan, Mr Lozos and Mr Simonson say: Designing letterforms means at the same time designing the white space within and inbetween them, so designing and spacing go hand in hand.
Igino says: "Who said that using iKern means designing without sidebearings in mind?"

Let me split up the discussion into two, to make it more clear that there is no contradiction at all. Rather there are two issues discussed at the same time:

(1) Designing type.
An experienced designer does work as Dan, Mr Lozos and Mr Simonson describe, he designs with both black and white parts in mind. And of course he would space letters as part of the design process, to determine the overall rhythm, and be able to evaluate the word image and check individual letterforms' integrity.

(2) Learning to design type.
Before even considering to use any automatism, one should have learned -- the hard way -- what spacing and kerning is about. And this is what Dan wrote: "The way to learn how to properly space a font is to keep designing and spacing new fonts." There is no way around it. One doesn't learn by being served results nor by reading one or two books about type design.

As to (1), Igino acknowledges this.
As to (2), Igino's concern is not teaching type design but spacing letters.

Relying on automatism or service -- what about human assistants? -- still requires that a designer knows what he does and wants. He must look at the work and criticise. Of course one cannot criticize without having trained the eye before which, again, requires practice, not just being served results or reading books.

In so far, I do not see any controversy here. The point that is of more interest to me is another one: That in type design praxis one may not get around making use of some kind of automation, in the long run.

iginomarini's picture

To James: You wrote: "What gives you that idea? You’re making an ad hoc assumption there about the way Dan Reynolds thinks, and we would all be better off if you did not try to read between the lines like that or put thoughts into people’s brains".

I just posed a question. Which thoughts are you talking about? Maybe are you reading between the lines and putting thoughts into my brain? “The way to properly space a font is to do it yourself” is a statement that can't be read between the lines by anyone speaking a language. I haven't found another way, except the question, to incorportate a statement posed in a postulate form inside a discussion, that is, as far as I know, to share experiences and renounce for a while to the absence of doubts.

Jos Buivenga's picture

Very well said, Karsten.

@ James: Please take into account that Igino very often has to defend his iKern. He's a nice guy.

dan_reynolds's picture

I think that iKern is an interesting service. Anyone has the right to use it. If they are happy with the results, I am happy that they are happy!

However, I do not see a benefit in my workflow of incorporating something of that external nature into the process.

If a type foundry asked for my recommendation (I work at a type foundry, although they have never asked for my opinion on this matter), I would not give a simple, one-word answer! On the other hand, I would express my skepticism. First of all, a large type foundry is unlikely to enjoy giving a typeface to space and/or kern to an individual using their own software. If in XX years, that foundry wanted to reproduce the workflow, and said designer/service were no longer in business, that would be bad. Second of all, many typefoundries are proud of their spacing and kerning habits, because they produce a sufficient level of quality across their product range.

Personally, as a typeface designer (and I am one, albeit a young one), I do not see the benefit in any method of spacing and kerning other than spending lots of time doing it yourself. This is my opinion, and is not meant as an attack on anyone else's.

k.l.'s picture

I do not understand the foundry-related arguments.
(2) Since kerning is considered part of design and thus left to type designers, a foundry-as-publisher does not have spacing/kerning habits, nor is it of relevance to a foundry whether the type designer spaced/kerned manually or not. And since spacing/kerning is done by the designer, it is him who pays for kerning (currency being hours) not the foundry -- in this respect a foundry does not even need to have an opinion about kerning methods. All that counts is quality of the result.
(1) The result of iKern is sidebearings plus a kern feature. Manual spacing/kerning also results in sidebearings plus a kern feature. Both Igino Marini and a type designer may go out of business and worse, as human beings, are likely to pass away one day. I never heard that the latter is a problem for any foundry. Once font data is there it is there, and nobody cares how it came into being.

William Berkson's picture

>Once font data is there it is there, and nobody cares how it came into being.

Small correction: font users indeed have no reason to care, but creators of new fonts may be very interested in techniques that worked well and those that save time!

dan_reynolds's picture

Oh, Karsten, I very respectfully disagree :(

(2) Having a designer carefully space and kern is a spacing/kerning habit. When foundries release designs produced by external designers, they still stand behind the data, it is part of the brand they work with. While each font in a foundry may have been designed by separate designers, the fonts are still produced and released uniformly. A foundry may have an "art director", who looks over designer submissions and makes suggestions for improvement. It has long been the case that many designers do not submit one round of completed files, but several different, improving versions. When a foundry or a library releases a new typeface design, the fonts represent a process between the designer and the foundry. Even if the foundry does not do the spacing and the kerning (and they shouldn't), they still do care how their fonts are spaced and kerned, and how they appear in print, onscreen, etc.

(1) It is always important how products you distribute come into being. Companies with traditions or with "long tails" very often go back into their archives. Perhaps a custom client wants to have a version of a typeface, with just slight variations, or more weights/styles/etc. It is good to be able to go back into the native files, and be aware of how things were produced, as well as to be able to repeat these results (or at least closely repeat them). Sure, you could re-kern a font so that it had the same values as an iKerned font. But if a foundry used iKern for all its new releases, and it then canged the character set of a font five years later, it would ideally want to send the new font data back to re-iKerned before its re-release. With one man operations, this is difficult to guarantee.

I should state that of these two points, number two would be more important to me if I were a foundry/distributor/reseller/etc. than number one.

k.l.'s picture

No worries about disagreeing. :)

(2) I am aware of the process you describe. And still the foundry is interested in the result -- quality standards to be met. The way in which a designer arrives at the result only becomes relevant if there is a need for improvement, then indeed one asks, and what is your approach, conceptually, that lead you to, uh, this?

(1) I do not see why, in your example, someone would want to entirely re-kern an expanded font to imitate previous kerning, if kerning already exists and only needs to be expanded. There simply is no difference between iKern output and manual output, in formal terms, both is a kern feature, i.e. kerning + class information. If you expand the glyph set, you'll end up doing some kerning additions, regardless of who/what the kerning in the first place.

Curioustype's picture

This is all terrific stuff; I especially enjoy and appreciate debates in which both sides have equally valid positions. To respond directly to a few things:

1.) Believe it or not, Mr. Marini, I am somewhat shy when it comes to one-on-one interaction of any kind. However, I can't say that's the entire reason I addressed the Ikern issue here. I'm not sure I know the main reason to be honest. If I had to guess though it would be this: Without knowing more of the fundamentals regarding your service, I'm afraid one of two things would likely happen. Either A.) I'd feel guilty for taking advantage of what seems to be a great service or B.) I'd get all excited about the whole thing and find out I couldn't afford it.

It would be different if I knew IKern might be released as software, at which point even if I couldn't afford it at first I'd at least be able to rearrange my already minimal finances and "splurge." I guess my whole point, then, is with more up-front information, many of my questions could be answered without the need to disclose any proprietary information. I'd suspect there are many people who might consider this service but would be embarrassed should they realize it was not affordable for them; and those same people might not feel comfortable with accepting any "service" for nothing.

2.) I thought this comment was profound: "This first time I felt I had to do it all to develop my eye; in future I would judge a computer-generated result better." I believe this is important as well because no one is saying designers are frothing at the mouth to completely eliminate the kerning process via automation. I think most if not all simply feel like they have a better chance at success, kerning-wise, if they are provided a "launching pad" so to speak and then make adjustments from there.

Besides, the bottom line is this: even for those who insist on doing things manually, I would think amidst a thousand kerning pairs there would have to be at least a few identical ones between your manual work and that of a program. And if that's the case, then how is it different to kern a typeface with X amount of space between characters (pre-kerned) and Y amount of space between characters (after automated kerning is applied)? Even if the latter was off, certainly it would seem to have pointed a designer in the right direction. I can't imagine IKern creating a +44 kerning figure to a pair you'd actually kern at -50. However if IKern generated a value of -26, wouldn't you still be closer to the final manually generated value than if you'd started at zero?

In the end it's all about analyzing - be it what IKern produces or what you do manually. I think if one were to fully understand the results IKern generates and adjust as needed - even if it requires adjusting every single potential pair - in the end you'll still have the kerning you want for your typeface. As for me, if obtaining the best kerning for a font required me to sing the national anthem and manually kern only the letter "G," I'd be at "OH say can you see," before anyone knew it.

James Arboghast's picture

@Ignio: I just posed a question. Which thoughts are you talking about? Maybe are you reading between the lines and putting thoughts into my brain?

Nope. You did not just "pose a question". It wasn't an innocent question. You phrased it in a leading way, leading to your own conclusion about how Dan thinks by way of suggestion and implication.

“The way to properly space a font is to do it yourself” is a statement that can’t be read between the lines by anyone speaking a language. I haven’t found another way, except the question, to incorportate a statement posed in a postulate form inside a discussion, that is, as far as I know, to share experiences and renounce for a while to the absence of doubts.

Come again? Can yopu put it in plainer English please?

@Jos: @ James: Please take into account that Igino very often has to defend his iKern. He’s a nice guy.

Too much an assumption of bad faith on my part, and more reading between the lines coming from both Jos and Ignio here. I never implied Ignio is not a nice guy. I only questioned his rhetorical stance and what *appears to be* an attempt to put thoughts into Dan's brain. I even took the time to explain that I have no views on IKern. I *did say as much*. I'm not taking anyone's side in this discussion. All I'm doing is defending one participant.

But a few of you are determined to see my intervention as some kind of attack.

The thing that would help most is to cease assuming bad faith. It gets none of us anywhere. I did not assume bad faith regarding Ignio's question for Dan. He asked a leading, loaded question, plain and simple. If Ignio can't see that or admit it, that's his problem, but it shouldn't be Dan's problem.

Do you people think you can handle that? Evidently you can't deal with it without abreacting, since the response has not been rational, and typically full of logic holes. I don't think I am asking or driving at a lot. It isn't exactly harsh criticism. I'm trying to reason with you guys, and it looks like at least one of you is determined to turn a polite discussion into a pissing match. I won't say who because personal remarks abot typophilers as people don't belong at typophile.

Ignio said: "Do you think that the type designers fully satisfied of my work are less competent than you?"

That's a leading, loaded question and quite suggestive. Hardly innocent. You literally asked if Dan thinks the type designers happy with his work are incompetent compared to himself (Dan).

Man, you people wouldn't last five minutes in my world (with the exception of Dan and Chris).

j a m e s

Bert Vanderveen's picture

James, don’t be so harsh — there are cultural factors in play here. If you’d transpose Ignio’s reactions into Italian you would understand there is no malice intended.

You, on the other hand, could be a bit more considerate, or one would have to classify your stance as typical American, which would be ironic, right?

. . .
Bert Vanderveen BNO

Jos Buivenga's picture

> I’m not taking anyone’s side in this discussion. All I’m doing is defending one participant.

Well ... that's exactly what *I* did. No reading between the lines.

dan_reynolds's picture

I'm not put off by any of Igino's comments about iKern. It is his product, and I am sure that he is proud of it. I am also sure that there are many customers who are happy with its results. I only lament the general discussion that there might be a way to simplify spacing and kerning in the typeface design process. In general, I have to say that I tend toward James Montabalno's comments in many of the previous thread about this topic. Thank you for trying to clarify things anyway, James.

What interests me most are typefaces. Not concepts or even processes, but typefaces… the end result. I like good typefaces. I think that most good typefaces are produced a certain way, and that there is certain evidence pointing to this, but hey! I could be wrong.

Bert, I wouldn't classify James or his responses as typically American. Although, I have to admit that I don't know what typically American is… and I'm not sure that there is anything typical about all Americans ;-)

Miss Tiffany's picture

James is Australian.

iginomarini's picture

I would like to say to everybody, and particularly to Dan, that I've never felt as being under attack. If there is an 'enigma' it is: why every time the word iKern appears on Typophile the same discussion raises? Since nobody questions the quality of its output, then is the context that have to be wrong. The metrics are in charge of the designer when he designs and also of the typographer when he set: these are the latest remarks. Every time cristicisms consist in traditional paradigmas applied to a process that's different and there's no way I could let envisage that difference. There is no discussion when someone says: I did so for years so this is the right way. It's just like writing posts in the critique section without having seen the fonts “discussed”. This is what I don't like. Anyway I'm the first to admit that my communication could be not enough. I've already begun to write new texts focused on these themes that, I'm sure, will make the picture clearer. Just wait. In the between I keep asking: test and judge yourself. If we were to find two parties in this thread they would be, in the end: those who didn't test iKern and those who did. Is this meaning something?

It seems to me that Dan essentially described Linotype that surely have one of the most, if not the most, wide and structured “production department”. On the other side there are one-man companies that probably don't sell licenses to institutions, banks or similar (for the moment). All the rest is in the between. I think that spacing and kerning are costs in any case. I can only say that to have, for example, an 800 glyphs font spaced and kerned in a few hours could make some difference somewhere. I'm not only referring to costs and timelines (that are money too). I'm referring to the fact that this possibility could shift the (market / dimensions structure) ratio and the (quality / dimensions structure) ratio to higher levels.

About iKern being a service: I don't think it will be a service the way it is today forever. iKern consists of two parts: an engine and a post processing module. The tuning of the engine really depends on the glyphs' shapes. The typographical choices are mainly done inside the post-processing module. I wouldn't be worried to give it away as a program so that designer could craft their own fonts by themself. Or maybe one day as a plugin for InDesign or Quark. Who knows? It will be the argument of one of my next texts.

AGL's picture

Following the info that was provided about iKern I figured Mr. Marini like to keep his trade secrets to himself and advertise his product, instead of showing a chart with the moving parts. Sounds like old school and it is his prerogative. Just email him if you need the service he provides. Looks pretty clear and simple. Imagine T. Edison talking widely about his light bulb, showing up the drawings, the pictures?


|\\\\\\\\\\\\| The Enigma Code Machine |\\\\\\\\\\\\|

Widely utilized during WWII to encode transmissions across the empire. As the encoded messages were, after encoded, transmitted via radio waves it was just a matter of time until the Allies cracked the hardware and the code, ending the atlantic and pacific wars. |\\\\\\\\\\\\|

William Berkson's picture

Igino, doesn't it make more sense as a plug-in for FontLab, or as alternate like FontMaster? From what you explain, the Font designer has a number of choices to make to 'tune' IKern, and also once the metrics are set hopefully the user of InDesign or Quark won't need to do much further kerning, if any.

Karsten, if you've tested both IKern and FontMaster, how do you compare the results and ease of use?

Sye's picture

@MissTiffany - typical australians... (i'm one too...)

blank's picture

…doesn’t it make more sense as a plug-in for FontLab, or as alternate like FontMaster?

Given a recent post to the ATypI list by a type designer whose Fontlab plugin is widely used and rarely paid for, it might not.

Bert Vanderveen's picture

[Software messing things up, sorry.]

. . .
Bert Vanderveen BNO

Bert Vanderveen's picture

James is Australian.

Ah, that explains it.

. . .
Bert Vanderveen BNO

Curioustype's picture

For the record, I think Mr. Marini is getting a bad rap here in some ways. It's almost like people believe Mr. Marini is developing the "Death Star" to use against the Rebel Typographic Forces on his way to conquering Earth. The truth is, whether anyone wants to believe it or not, he might be sitting on something quite lucrative before all is said and done.

However, to respond to this statement: "Imagine T. Edison talking widely about his light bulb, showing up the drawings, the pictures?"

If my history is correct (and good luck to me on that), Edison never made a dime on the light bulb, or anything else he invented, instead choosing to name it something like a necessity for humanity.

Also, I wasn't particularly concerned with the inner-workings of IKern, just the fundamentals. Sorta like, "This a light bulb and it creates light through electricity. Though in the test stages, we expect it to be released for general consumption by the public in X years/months, at an approximate cost of $X.XX." Or, "we plan to provide light bulbs to everyone who'd like to have one before the end of the year 1903, at a cost of $X.XX per month, or X.XX per usage hour." As it is now, all that's being said is, "We've got this thing, a light bulb, that will light your home like nothing before it. We'd be glad to test one in your home," without stating what comes after the test, whether or not it will be widely available or merely rented to those with the means to afford it, etc.

I'm not necessarily complaining about the lack of information, I just live in America where people come out of the womb suspicious of the big guy with the blue mask and shiny tools in his hand. "Hey, whaddya planning to do with that thing??"

William Berkson's picture

For what it's worth, I suspect that Igino Marini has "built a better mousetrap", but that selling it as an application is a better way to go. But then I certainly have no expertise on marketing, so I could well be wrong. I am pretty sure that it will be easier to sell an application, rather than a personal service, to type designers. But how the math--and money--would work out in the long run for Mr. Marini I have no idea.

acnapyx's picture

I already wrote an answer in the beginning of this thread, and while considering if I should post it (it was quite in the harsh tone of Mr. Arboghast), finally decided not to. After all, I did not want to dampen anyone's reputation (especially Igino's, whose name in fact was so often misspelled in this thread that it sounds really humiliating).

But - just for the matter of discussion: you can in fact compare the results of iKern and KernMaster in quite simple way: take a typeface that is known to be kerned with iKern; save the kerning to AFM and strip the values to have a set of kern pairs to load into KernMaster. Then try to kern the same font. With same set of kern pairs I think you can judge which one is better.

I have to admit Fell Types are not the best examples to use, but the results with same list of kern pairs differ substantially. I have done this comparison quite long ago, and I haven't kept the results, but sometimes KernMaster offers negative values where iKern has calculated positive ones for the same pair. So, I trust the intentions of the developers of such algorythms, but no software is perfect.

Personally I do not feel comfortable outsourcing kerning work, and although it is a tedious task, I often rewrite so much the KernMaster results that I ask myself why I attempted to automate the kerning in the first place. Of course I heavily suck in spacing and kerning, so I need a crutch to get the thing halfway, but I do not rely on such software that much, so I still do the spacing/kerning by hand - not least to improve my skills.

Personally I would prefer seeing iKern at least as a web-based service, served by scripts on a webserver, if Igino does not want to distribute it as a software. I would prefer to do the things myself, alter some settings and see the results in real-time. But after all, this is Igino's work, so the decision about how it should be used is left to him. If iKern was offered as a software, surely I would try to scrape my savings and 'splurge' (hope this happens someday). Until then, I'm not that much interested - just curious.

Miss Tiffany's picture

I wasn't trying to say anything specific about Australians. I was only making a correction. :^P

I think it is unfortunate that people have to be so negative and can't just discuss things in a way that promotes education.

dezcom's picture

Igino Marini deserves credit for attempting this effort. I have not tried it or even spoken to him but I would be curious to see how his service works.
I don't think anyone here is trying to insult Igino, I just think we are a curious bunch and many of us have been bitten by anything that sounds like it is automatic or says it is time-saving. Anyone who has ever tried FontLab's auto-kerning can tell you it is a nightmare to fix the mess that it makes and you are better off staying away from it. I assume iKern must be much better than that if Jos praises it.
We have to distinguish the difference between learning what good spacing and kerning is (along with becoming skilled in our craft) and having a time-saving tool that produces respectable work. The proof is in the pudding but most of us have yet to test it.

To Igino I will add that if anything I said was taken by him as derogatory or insulting, I am sorry and I assure you it was not intended. I wish you great success in your work.


Jos Buivenga's picture

A short story about how I got to know iKern ... For some years I was looking for a tool that could do a good auto-kerning job. I first came across iKern via Typophile, but because Igino's previous email address was heavely spammed I didn't get a reply. In March 2008 I read the"from Moleskine to market" article on iLT from Kris Sowerby where he mentioned he did his kerning with the help of Metrics Machine from Tal Lemings. I tried to contact Type Supply, but again didn't receive a response. Karstens site put me on the trail of DTL Kernmaster, which I thought I should keep i mind because it looked like a great option. When I was working on a new version of Anivers Igino contacted me via my blog and I recieved a reworked version of my earlier Anivers regular. I wasn't aware at that moment that iKern could also do auto-spacing. Igino was able to capture the spirit of Anivers with a spacing and kerning that made it sing. That really convinced me and we decided to collaborate on the iKerning of the complete Anivers family. In the process different versions were send to and from and always I felt I was in control of the workflow. You can *test* the free regular end result for yourself. After that Fertigo Pro and Museo Sans followed.

iginomarini's picture

At this point I think a clarification is needed. Every discussion on iKern as a marketing model, are out of place from the beginning because simply there is no a planned marketing model. iKern is not my job and is not my living. I see Jos has talked about how we first met on Anivers. He hasn't talked about his efforts to convince me to be paid because otherwise he would have felt like stealing. This is my marketing model! So if I had to spend six years developing a project to make some money being this project an autometrics tool then I wouldn't be that smart. If I wanted more money I would have worked more. Just simply. I spent a lot of time on iKern because I don't sleep too much! But, mainly, because it satisfied an inner exigence. I've conjugated an ideal of beauty, simplicity, essentiality, lightness and truth with logic, method and complexity. My arid talents. I've found my personal way to create something. It still makes me somewhat happy. This is the real reason. What's more I'm also an engineer. That's why I had to make something that works well. Having this thing in my hands it seemed natural to me to propose type designers to enjoy. I thank Chris for his wish of success. My success already is when the iKern's output is defined “fantastic”, “amazing” and so on by everyone used it. My success is when someone thanks me for having solved his problem. My real satisfaction is when Jos says: “Igino was able to capture the spirit of Anivers with a spacing and kerning that made it sing”.
I have already said that today iKern is a service. I'm imagining how it will be tomorrow. Today I feel comfortable this way for many reasons not being the last that it's still a work in progress and that I wouldn't want to be in the position of having to defend my job because of others' errors. Not in this phase when I have still a lot to do defend my job itself! I still think the service it's a good way. On my site I've tried to explain, and I will do more, why it's not a conceptual error to conceive an intermediate figure between the type designer and the typographer. Type designers and typographers permitting! About the costs. I have to say it's not me who decided my basic fee. So I'm confident it's correct. I say basic fee because I keep the freedom to decide case by case. Let's say it's the equivalent of three hours of work: so everyone spending more than three hours spacing and kerning a font should be glad.

Sye's picture

Hi Igino and Jos, one question, does the type designer do much spacing themselves before handing the files over to iKern, or does the iKern service look after that?

and to those who have been against this initiative, how is it different to hiring a 'spaceing and kerning specialist'? I'm pretty sure i've heard it's not uncommon for some type designers to have their files optimised by others (usually from the foundry.

All this talk of spacing a kerning makes me wonder why Mrs Eaves has not been re-spaced for text settings, does anyone know why?

Jos Buivenga's picture

Yes, I do space myself, because it is part of the design, but much more global than before. Often I ask Igino for an iKern pre-run when I've determined how I want most characters to be spaced.

Syndicate content Syndicate content