It's because you are not used to FontLab. Illustrator has great drawing tools, but for letters, FontLab works better, once you are used to it. Besides, the biggest pain point for drawing fonts in Illustrator is getting the letters into FontLab. Drawing in FontLab solves that problem beautifully ;)

Eben,

I do want to embrace the learning curve in Fontlab but I'm finding it difficult to do what would be really simple in Illustrator.

In a post form early today I was asking about aligning two outlines, in that case it was the slash form the Oslash. In this situation is it a case of having to center the slash by eye or is there a method of centering the slash on top the O gylph.

The soultion I was using involved placing the slash as a seperate glyph and then adding that glyph as a compent which I could then center, then decompose and finally deleting the slash gylph. I'm sure there must be a better method than this.

Rr

crossgrove,

I understand I need to leave Illustrator and start drawing in Fontlab, and I understand that fontlab containes everything I need, but I'm just struggling with some minor points.

The example I talked about using composites wasn't creating glyphs with accents as it would normally be used, but it was a work around I devised to try and solve the problem of centering the slash in the Oslash glyph.

Everyone ha sbeen really helpful here but I still carn't work out if Fontlab has a method of centering such things or if it's best doing by eye.

http://typophile.com/node/34540

Rr

Sometimes the reason things are different in FL won't be clear and then you will realize 'why' later & sometimes Fontlab may just suck in one way or another. The reality overall doesn't change though.

Moreover, while FL may have a tool for centering things, even if it does you may well be better off doing it by eye. Mathematically correct is often not right when it comes to the eye.

For more on this see this:

http://briem.ismennt.is/

But especially this

http://briem.ismennt.is/2/2.3.2a/2.3.2.06.illusion.htm

With regard to the mathematical vs optical, the O and o slash actually are good examples. Note that they in effect have two of the 'x' style illusions in them of the kind that Briem mentions. As Mark Jamra notes here, to make it look 'right' you may want to rotate the middle line relative to the others. In this case what 'centers' can only be done by eye, particularly if your O and o are not vertically symmetrical.

Rachel,

Precisely...
i´ve also posted that question some days ago about the same issue concernig a method to align shapes in FL without having to compose/decompose glyphs.
Nobody replyed though...
I know opticaly is the way to do it, however aligning mathematical could be a good start point.

Rachel:

I was like you the first time I launched FL to draw letters. I was used to work with Illustratror since is first version (88). I did some switch in the beginning (draw in Illustrator, import in FL, go back and forth between those two of them...). However, I learn FL and his tools, with the Cabarga manual and now I find FL is very in first place and works better to draw letters.

Just thought I would come back and say - yes Fontlab is better than illustrator for drawing letters. Its only been a few days and I'm still working stuff out but I've saved a ton of time already.

Wish I'd made the move sooner.

Thanks for all your response.

Rr

Rachel,

Your recent message is the best post in the thread! Congratulations. You work quickly!

Not-so-comforting thought: FL has some severe bugs and non-intuitive functions, but overall I marvel at the power it has, for things I always wanted, like putting an entire font in the background, and building all my composites in a single operation. It's deep like Photoshop is. And I do expect FL to improve; the programmers are very nice and very talented and have been absorbing a lot of feedback since Studio 5 came out.

Russ,

The time saved is not drawing in Illustartor and then having to copy into FL, it dosen't sound like much but when I'm copying 200+ glyphs, the time adds up. The points in fonlab will only snap to the grid (I hope so anyway), but in Illustrator even with snap to grid on and the grid set up correctly I found that some points will move when copied into FL. I found it really frustrating that I spent the time drawing in Illustrator to then have something slightly differnt in FL.

I've only just started drawing in FL, so I'm no expert but I found it really simple to pick up. I think I've finally realized that drawing in FL cuts out the middle man. Crossgroves first post on this thread sums it up for me.

Rr

Leslie Cabarga explains that the basic difference between Font drawing programs and Illustrator is that when you delete a point, the line doesn't collapse in, but springs out. This is useful for shaping curves, as the 'sprung' curve will have a simpler geometry, with no lumps. You can then decide whether you want the extra node or not.

> I'm pretty good at drawing curves accurately the first time

But there's no way you're pixel-accurate.

Using Illustrator to draw glyphs is virtually never a good idea.

hhp

I should have written em-accurate. What I mean is, you can't
for example make the 3 stems of an "m" the exact same width,
right off the bat, freehand.

hhp

One more thing. FL is probably much better when you do the outlines point by point. But I normally draw glyphs using lots of welding/trimming/intersecting. I can reposition each part separately and then do the weld on all of them and do the final tricks on the outline itself.

So when it comes to some geometric sans serif it's rather easy to take one part of a letter and use it with another. I don't really know if this would be the case with FL. But if you draw your letters on paper so good you scan them and do the outlines in FL that's as you all agree much better and easier with FL.

> with [...] mathematical precision

My background is in mathematics, so I will jump in. If the counter is perfectly circular then there is a center. If not, the counter has a "center of gravity" i.e. a point such that if you cut the counter in straight line through it, both parts have the same area, the same "amount of white". If you select a few points on the boundary of the counter and get their center of gravity, it won't be that of the counter. Moreover, even if you have a program that gives you with mathematical precision the center of gravity of the counter, as soon as you draw a thick line through it, the two white pieces in general stop having the same area, i.e. the same amount of white and I understand (please correct me if I am wrong) that what is important is that the amount of white under and below the slash be the same. Moreover, other issues due to the Poggendorff optical illusion need to be taken into account. So the question is not a simple mathematical problem. If we forget about the necessity to correct the optical illusion, then we would need a program able to calculate the area above and under the (thick) slash; is there such a thing on the market?

"(please correct me if I am wrong) that what is important is that the amount of white under and below the slash be the same"
Okay.
The amount of white below the slash is greater, slightly.
This is what made the fog method perfect.
The slash was centered on the inner counter,
and then nudged to visual certainty.
Cheers!

>I have lived all my life with eights having an upper part that looks smaller than the lower part...

In general whenever a character is divide in two, it looks more 'right' when the top is smaller than the bottom. Partly, there is the visual illusion Briem notes where eg the bar of the H (or the O slash) has to be higher than the center to make the top and bottom look equal. But more than this I think we react as if the letter is an object with weight, and things are more physically stable when the top looks smaller, not equal. Whatever the reason this holds with many horizontally divided characters, such as the B, S etc. Visual equality is what you may want in an H or O slash, but often not in an E and 8, etc. In almost every character, the mathematically simplest shape and proportion looks awkward or ungainly.

> But more than this I think we react as if the letter is an object with weight

This is also the way I perceive it personally.

>But critically, not during immersion.

You have a good point, but I think 'aesthetically' rather than 'critically' is more accurate.

Readers don't generally set out to criticize a type face. They normally take it totally for granted. But they are strongly affected by aesthetics during the first glance at a page, and particularly the titles, headings, and layout. Then when they read extended text the face should 'disappear' as they focus on meaning. During orienting glances is when aesthetically pleasing shapes and proportions have their impact.

>kind of thing we’re making, like a text face vs a display face.

I guess here is where we have some difference of opinion.

Unless a face is intended for tiny sizes only, like an agate face, text faces also have an element of 'display'. Even if they are at text sizes, the aesthetic aspect, including proportions, has an impact each time we have a first look at a page, or perhaps even a paragraph. I agree that the beauty of type at heading and titling sizes is more important. If the face is done in optical sizes, then the heading and titling sizes can successfully put a bit more priority on looks and a bit less on readability.

Well, if that's what you meant we're in agreement.

> For those who prefer to draw in Illustrator I highly recommend to use BitFonter or ScanFont to transfer AI drawings to FontLab Studio.

Can you expand on this?

Just out of curiosity, if one is going to use Illustrator rather that Fontlab to create the glyphs of a font, can anyone list the steps that should be taken in ensuring the imported vector art will adapt to Fontlabs specs? What needs to be double checked or updated - ie. redundant beziers...

sb

I just came across this link in another thread.

http://groups.msn.com/fontlab/tipsandtricks.msnw?action=get_message&mvie...

anyone have anything to add? or does this pretty much cover what needs to be done?

Drawing letters in Illustrator is very simple if you have lots of practice. I've been doing it since 1989 on a Mac of little substance. If you have the "eye" for it, you need not use many beziers, only the minimum necessary.

I have worked with artists who use many to quite a few bezier points, so I know what many of you might be suffering through. I would love to give lessons on the minimalization of the bezier-drawn letter. It is so easy once learned.

Having said that, how well would a letter with the minimum number of beziers transfer to Fontlabs? I am interested in creating some of my own ideas.