FontArk

Hi

Did anyone here try Fontark?

I'm involved in the development of it and would like to hear (and share) some opinions, it is still in early stage but already capable of some wonders.
We represent a unique glyphs synchronisation system (The SmartX system), which is most versatile and aimed to replace the "scissors&paper" like actions of slice-cut-paste-merge shapes with real-time multiple character editing.
It's mostly demonstrated with Skeleton usage, but works cool also with strict outline drawing. It's principles and operation are most simple and very powerful, yet challenging to explain in an elevator .

I'd love to demonstrate it in action if anyone's interested, and answer any question.
«13456

Comments

  • Chrome or Safari only? No Firefox?
  • Ofir ShavitOfir Shavit Posts: 289
    No Firefox Patrick
  • It is too lacking in the basics to really test it well: undo/redo/nudging points/selecting multiple points etc.
  • Ofir ShavitOfir Shavit Posts: 289
    @Frode Understood. We've inherited the limited vector editing tools from Svg-edit, but we're gaping up, it's important to know where to focus our development efforts, thanks!
  • Ofir ShavitOfir Shavit Posts: 289
    @Frode The nudging points + selecting multiple points has been implemented a week or two ago (Undo/Redo was there from the start in these arrows image image in the bottom of the work area).

    Of course there are still missing features but most of what needed for designing a font with it is there, the most important thing though is to understand how it works, I mean...you only draw the Skeleton, and connect it to the Smart-points (SX). understand how and when to modify a single glyph, several glyphs together and the entire font, from there on it's pretty intuitive.
  • Ofir ShavitOfir Shavit Posts: 289
    edited April 2014
  • Ofir ShavitOfir Shavit Posts: 289
    Rob, In this case the end result has nothing to do with the tool's capabilities, especially the important ones.
    It'll be like looking at cars at the end of the race... great, they all made it to the cross line.
    (except one is a Koenigsegg and the other is a Lada). Or if you'd like, looking at a font and guessing whether it was made with Fontlab, Glyphs, FF or any other font editor. That would be a total missing the point.

    Fontark is a font editor, or a font creation tool, still in development.
    It is designed for full freedom of creation and you got there almost all the standard vector editing tools you're all familiar with, and if you know what you wish to achieve you can do it with Fontark as well.

    But Fontark is not like any other font editor. underneath the vector editing layer we've built up a unique and very special glyph synchronization system, which for the first time (as far as I know) gives a type designer, with simple graphic tools (no coding no geeking around), the ability to freely construct a set of characters over a fully controlled synchronization system that let him perform a great deal amount of design operations over the entire character set, whenever he finds it appropriate, or over a selected group of characters, or a single one. operations which normally he has to imply manually one by one to each and every glyph around.

    It's not an app, it's not a tweaking tool or an esoteric gadget, it's full free-design platform.

    So in this case it's all about how you get to the end result and not the result itself. There are some steps in the way which we ease a lot and some left untouched. We surely have some missing functions as well at this stage. I'd love to talk with you about those missing but I think there's quite a lot to talk about the progress we made so far and it's merits and demerits.

    But you have to get a certain picture of the guts of the tool, otherwise you're just running in your mind with i don't know what that is probably irrelevant (to the essence of what we're doing, to the stage of our development, it's potential and to how it might influence the "world" etc')

    The core issues as I see it are:

    1. How are the characters synced?
    2. How can you control the synchronization?

    From these two you can figure out:

    3. What is it good for?
    4. How is that making things easier?

    The easiest way to answer these at the moment imo is by a simple demonstration but let's do it your way. I'll be posting here soon some information and what you've asked for and see where we're getting :)




  • > ...which for the first time (as far as I know) gives a type designer, with simple graphic tools (no coding no geeking around), the ability to freely construct a set of characters over a fully controlled synchronization system that let him perform a great deal amount of design operations over the entire character set, whenever he finds it appropriate, or over a selected group of characters, or a single one. operations which normally he has to imply manually one by one to each and every glyph around.

    Actually, there were a bunch of font compression systems back in the 90s that had corresponding proprietary (in the sense of, not licensed to the general public) editors that apparently worked this way. The only one I got to use was ChamEdit, the editor for the Ares FontChameleon technology. As far as I recall, the editor really didn't require any coding or geeking around, as you put it, unless it was just entering in a handful of global values for measurements in the font. You could prototype a typeface in an hour with that thing.

    Of course, there were differences of opinion on how useful and interesting the results were. I was about the only one on the Adobe type team who was excited about the utility of this as a prototyping tool, as best I recall.
  • Chris LozosChris Lozos Posts: 923
    As a previous owner of Font Chameleon, I can see what you mean, Thomas. This is what I referred to when I previously posted "smooshing fonts together."
  • Ofir ShavitOfir Shavit Posts: 289
    That's why I say you have to get some picture of Fontark's guts (engine). It has nothing to do with blending fonts and manipulating existing fonts.

    Fontark is a blank canvas.

    The users Draw the characters into it, with drawing tools.

    The connection of the characters is defined by the user by simply connecting the Nodes of the paths to special cross-glyphs smart points. (which are also controllable).

    You can draw any shape and connect it anyway you wish with any other shape (that's why you can use Fontark's platform to create icon fonts and logotypes as well and not be restricted to any programmed scheme)

    Once you have a set of connected characters/shapes you can keep on shaping them freely only now, instead of repeating your path editing on each character it'll happen simultaneously according to the way you've connected the glyphs... entirely with graphic tools. The only slider bar control we have is the outline-width which control the weight of the font.

    So fontark is designed first and foremost as a Design platform, nothing else. and you can design with it this this this and this and any other thing on your mind, in the most simple intuitive and flexible way I know.



  • Michael ClarkMichael Clark Posts: 129
    edited June 2014
    It is etch-a sketch for font designers. Now we have arrived at higher thinking. God bless us all!
  • Ofir ShavitOfir Shavit Posts: 289
    Ok, @david Berlow

    You wanted a
    complete BL set of skeletons covering basic Latin for regular and italic postures of some range of weights and widths of your sans
    I thought about starting with this font, Xumps, specially made for the occasion.
    It's quite basic but not the easiest I could think of.

    Before I start making the range of weights and italics (for each weight right?) and several other variations I'd find interesting, Do you have any objection/remark on that font? if you do please suggest what you have in mind.

    Later on I'll upload it to Fontark as template and you could all see how it's built (with the skeleton) and play around with it if you'd like

    Few remarks on my side before we start:

    1. try to leave aside design issues (I'd love to hear any but outside of the context of this demonstration which is mainly technical as I see it)
    2. This font has no Optical corrections, Kerning, Hinting and some other features we didn't yet deal with in Fontark
    3. If you (anyone) wish to challenge me with tasks regarding this font feel free to suggest any, I'll do my best to follow your challenge and explain what I do and how (in case it's possible of course)
    4. the '@' is missing. Known issue.

    Attached you'll find Xumps in otf, raw as downloaded from Fontark (plus some name ids edited with my OTMaster, thanks to Jan, Frank and Mark) and an image of Xumps as text

    P.s. If anyone can tell me why the heck in Inkscape I get the character 'f' instead of 'space' I'd be thankful! :)
  • Ofir: ‘[…] and you can design with it this this this and this and any other thing on your mind […]

    I think that you perfectly formulated here one of the biggest criticisms from type designers: if something has been defined, one can make a serious attempt to artificially reproduce it, but is it possible to artificially define something completely new from scratch? So far all attempts I know of, including yours and mine, will have a long way to go before this is possible -if ever.

    March 2013 in a discussion on parametrized type design on the ATypI list I referred to the
    Experiments in Musical Intelligence (EMI [pronounced as ‘Emmy’]) project by David Cope, who is professor emeritus at the University of California. The EMI Project started in 1981 and can be described as a method to come to new compositions via reusing parts of existing pieces of music, also referred to ars combinatoria. This method requires that the EMI program can analyze and decipher existing compositions.

    A project like EMI is not welcomed by everyone, because it relativists mankind’s creative side. Sometimes negative opinions are adjusted though. The American scientist Douglas Hofstadter, especially famous for his book Gödel, Escher, Bach: An Eternal Golden Braid (1979), describes in the article Sounds Like Bach his initial demurs concerning computer-generated music: ‘Back when I was young – when I wrote Gödel, Escher, Bach – I asked myself the question “Will a computer program ever write beautiful music?”, and then proceeded to speculate as follows: “There will be no new kinds of beauty turned up for a long time by computer music-composing programs… To think – and I have heard this suggested – that we might soon be able to command a preprogrammed massproduced mail-order twenty-dollar desk-model “music box” to bring forth from its sterile circuitry pieces which Chopin or Bach might have written had they lived longer is a grotesque and shameful misestimation of the depth of the human spirit. […].

    A quarter of a century later Hofstadter played a mazurka generated by the EMI program in the style of Chopin, who he admires: ‘[…] Though I felt there were a few little glitches here and there, I was impressed, for the piece seemed to “express” something. Had I been told it had been written by a human, I would have had no doubts about its expressiveness […].

    John Hudson posted an interesting remark: ‘The example of a mazurka à la Chopin is interesting, because it involves both a set form or genre with built in structural constraints (the mazurka) and patterns of idiom (Chopin's individual style), both of which can be captured by analysis [...]. You can analyse Picasso’s paintings of a particular period all you like, and easily produce pastiches of them, but no amount of that analysis is likely to make it possible to predict what he will do next.

    Cope seems to refer a bit to this in an interview from 2002/2003 when he mentions ‘freely’ composed work: ‘Music composition consists of a combination of what we hear and what formalisms we bring to bear. If I compose a work freely (i.e., without a prescription for voice-leading, allowable verticalities, etc.) then I will most likely integrate various ideas that I’ve previously heard. If I compose a piece strictly using a mathematical formula, then I won’t be re-composing music that I’ve heard but following strict rules. Most music consists of a combination of these factors. The notion that humans have some kind of mystical connection with their soul or God, and so on, allowing them to produce wholly original ideas (not the result of recombination or formalisms) seems ridiculous to me.

    I must admit that deep in my heart I want to disagree with Cope, although he has a good point. But so has John.
  • D. Epar tedD. Epar ted Posts: 673
    "Do you have any objection/remark on that font?"

    Yes. For the purposes of this demo, it is not advantageous to have an "extreme" design in any one of the following areas: overly square, extremely high-waisted, low legibility proportions or extra doodad details.

    So, for this exercise, make it rounder, even at the cost of looking a lot less unique.

    And for this exercise, visually center the crossbars of all the uppercase vertically.

    And then, for this exercise,
    balance the f and t so the left bar isn't a stub,
    and shorten the t ascent,
    flatten the e's crossbar,
    remove the 'serifs' from the k, x and z
    decide on one tail shape for l.c. l, and t.
    end all the diagonals with flat cuts, or not
    and fix the W, w, so it's related to Vv and Yy.
    before we decide what to do with the g?

    I would not like to see what happens to the weight and size variations otherwise.
  • Ofir ShavitOfir Shavit Posts: 289
    edited June 2014
    Frank, really, this has nothing to do with Fontark. Fontark is MUCH simpler and analogue than that, and in no way a step toward letting stupid machines make decisions for humans, on the contrary, Fontark is a step toward efficiently making a computer do what it's good at = doing what it is told to!
    It's as if you (a conscious being!) had 100 hands instead of one. hands, not brains.
    Controlling 100 hands obviously requires a slightly different attention (than focusing on one hand at a time) but when you understand how simply Fontark works and how easily you "turn off" a hand here and a finger there from moving in an unwanted direction you find it most natural.

    David, No probs. let me arrange these.
  • Ofir ShavitOfir Shavit Posts: 289
    It is etch-a sketch for font designers.
    Want to have an eat that hat bet on this?
    :)
  • Ofir ShavitOfir Shavit Posts: 289

    How's this?
  • Ofir ShavitOfir Shavit Posts: 289
    edited June 2014
    Just to give a sense of how the changes are done...

    http://t.co/dE1FHMEYnX

    — FontArk (@FontArk) June 15, 2014
  • Ofir ShavitOfir Shavit Posts: 289
    edited June 2014
    BTW Frank, imho Cope is only partly right, his mistake is that he takes out of his clever equations the elements he is unaware of (and or have no proofs for) and believes the results of his (in a way, ignorant and naive) calculations. John is absolutely right, and much wiser, understanding that we know less than we don't.
    Inspiration arrives at the present. What manifests in the mind of a person is the reaction of external influences of the moment with the specific structure of his current psycho-physics and all he has been through up to that point.
    The psycho-physic complex of a person at a certain moment is obviously not describable but I think that it's the external influences that might be most surprising and those we're most ignorant about, including possible Ideas, lines of thinking, emotional states and rhythms that might get caught from the external at a certain frequency of attention. after all, dummy cellphones easily catch vibes of something allowing transferring messages. The human organ is much sensitive.
    Or in short, don't get attached to your thoughts, they might not be yours in first place ;)

    As for Fontark, I don't know where the ideas for it come from but I can assure you it's being uncompromisingly submitted to the service of the human spirit.
  • Ben BlomBen Blom Posts: 161
    edited June 2014
    David Cope: “The notion that humans have some kind of mystical connection with their soul or God, and so on, allowing them to produce wholly original ideas (not the result of recombination or formalisms) seems ridiculous to me.”

    This fits in with how I would define designing and creativity. Designing is deciding. Designing is problem solving. Creativity is the art of combining existing components in a new way.

    Henri Poincaré: “Invention consists in avoiding the constructing of useless combinations and in constructing the useful combinations which are in infinite minority. To invent is to discern, to choose.”

    Frank Blokland: “I think that you perfectly formulated here one of the biggest criticisms from type designers: if something has been defined, one can make a serious attempt to artificially reproduce it, but is it possible to artificially define something completely new from scratch? So far all attempts I know of, including yours and mine, will have a long way to go before this is possible -if ever.”

    If creativity is combining existing components, it will never be possible to define (create) something completely new from scratch — artificially or in any other way.
  • D. Epar tedD. Epar ted Posts: 673
    '..make it rounder,'

    "How's this?"

    I said Rounder, meaning more round than rectangular. This introduces a group of parameters you'll not face otherwise, called spaces.
  • Ofir ShavitOfir Shavit Posts: 289

    Spaces are fully configurable with us.

    1 or 2?
    1-2.jpg 239.8K
  • Robin MientjesRobin Mientjes Posts: 92
    edited June 2014
    At the risk of being obvious, I think it’s mathematically possible to decide which of these is the roundest.

    To balance for the sass: David means you should go round because one of the leading things in spacing a typeface is the spacing of unequal sides. If every letter has straight sides, the equation becomes a lot simpler. But real-world complexities exist when you have straights, rounds, diagonals and the occasional ‘I am not sure how to qualify this shape’. Straight-side everything makes it harder to judge the capabilities of your tool, in short.
  • Ofir ShavitOfir Shavit Posts: 289

    I've told it before, the capabilities of Fontark (in this case for sure) are the capabilities of it's operator. Fontark will not space the letters for you, it's not an AI thingy (thanks lord). Instead, it helps you do it with what we call... Real time multiple glyph editing approach.
    Instead of manual one by one editing or accountants classes defining you'll be selecting the glyphs you want to modify and modify them together.
    If you won't blame Fontlab for badly spaced font made with it, you shouldn't blame Fontark for such thing as well. You might prefer one method over the other though.

    If you're interested with (total) round I'll switch to classic font structure
  • Chris Lozos wrote: “As a previous owner of Font Chameleon, I can see what you mean, Thomas. This is what I referred to when I previously posted "smooshing fonts together.”

    Ofir replied: “That's why I say you have to get some picture of Fontark's guts (engine). It has nothing to do with blending fonts and manipulating existing fonts.”

    Urm, fellows, that is not at all what I was trying to talk about. I am sorry I was not being clear enough! I take full responsibility.

    FontChameleon was, among other things, a retail product. People could use it to take the built-in “font descriptors,” and then for any one of them, manipulate the x-height, weight, width, and other parameters. Less excitingly for type design, you could also “blend” these descriptors, which was an amusing toy, but often not very useful.

    The ChamEdit app was used to create these font descriptors. It was not a retail product, but a proprietary Ares tool. The amazing thing was, you could create a basic descriptor in about an hour, getting a full western+CE character set. The descriptor could then be used to generate fonts of varying weight, width, etc.

    I was just really impressed with this as a tool for rapid prototyping of some kinds of relatively conventional typeface designs. It wouldn't work for script fonts or dingbats or anything too bizarre. But it was remarkably powerful, and quick to use.

    (Of course, Adobe bought Ares for different reasons: the descriptors were very small files indeed, and offered a form of radically small font compression. So the PostScript/printer group at Adobe drove the acquisition, nothing to do with the font team really.)
  • Ofir ShavitOfir Shavit Posts: 289

    Thomas, what were those 'descriptors'?

    We've just published a useful new feature...

    Optimal corner curve - Handy new Feature. http://t.co/13CuezRMoE

    — FontArk (@FontArk) June 16, 2014

    (I'm embeding here the Twitter twits because it brings with it the video clip. If you know any other way to embed videos directly to the post here I'd love to know it and stop that habit :)

    Could you do things like this with ChamEdit?
  • Nope. But remember, ChamEdit was 15–20 years ago. It was pretty amazing at the time. :)
  • Chris LozosChris Lozos Posts: 923
    Why is a perfect circle the "Optimal Curve"? That is really a less than stellar S you've drawn and it wasn't even a sensible process to draw it.
  • Ofir ShavitOfir Shavit Posts: 289
    Chris, We're having hard time naming this feature.

    First, this feature refers the action that is being implied at 0:35, 0:40 and 2:12 in the clip.
    All the stretching and scaling in the clip just demonstrates this feature's usefulness.
    It is not a circle feature but it is based on cubic bezier circles math. It'll result with a perfect circle when you position the two segment's Nodes on a 45deg axis but you can apply it to any curve segment, in most cases you'll get a quarter of an ellipse.

    Optimal corner curve is the closest I could come to describe this feature, I'm open to suggestions:)

Sign In or Register to comment.