Interpolating diagonal curve endings

Frode Bo Helland's picture

How do you deal with diagonal curve endings when interpolating in Fontlab? I’ve not yet figured out a good way to handle them. I’m using the method describe here, but it offers no advice on this issue.

blank's picture

I’ve stopped using the interpolation tool and work with Christian Robertson’s Interpolated Nudge script and Göran Söderström’s macro tool script to let me drag points into place. This allows me to just sculpt cuves into place without worrying about the limitations of the interpolation tool.

Frode Bo Helland's picture

Would you care to share a link?

Frode Bo Helland's picture

Thanks a lot!

Quincunx's picture

That nudge script is pretty handy. Thanks for the tip. :)

mk2's picture

Great tip, Mr. Puckett.

blank's picture

If you’ve never spent a few days testing out every Fontlab script on the internet, trust me, it will save you time in the long run.

Mark Simonson's picture

I can't believe I didn't know about Christian's nudge script (and Eigi's tool version). Very handy. I need to get out more.

Quincunx's picture

> If you’ve never spent a few days testing out every Fontlab script on the internet, trust me, it will save you time in the long run.

Well, if you have any more tips... please do post them. :)

blank's picture

The neighbors feature is much more useful if you replace the default config with a file that displays your control characters around letters.

Frode Bo Helland's picture

I’m slow:) I didn’t try this out until now, and I can’t seem to get it working. Moving a node does not effect the control points at all. Any tips?

Frode Bo Helland's picture

Never mind, I got it working. It however does not help me the least controlling angled terminals any better.

John Hudson's picture

James: This allows me to just sculpt cuves into place without worrying about the limitations of the interpolation tool.

Can you explain to me what you see as the limitations of the interpolation tool. I use this tool a lot, have done for years, and I'm always surprised when I hear people say they don't like it or find it problematic.

blank's picture

In my mine the limitations of the interpolations tool for are:
• The inability to change my zoom once I start using it.
• The nature of the tool is the alter the entire letter unless I lock down a bunch of points.
• I have to repeatedly drag points to different potential locations until I get a result I like.

These items it very inefficient for shaping letters when I do not have a clear idea of what the end result should be. It is still a great tool for quickly adding or removing weight, creating small capitals, etc.

John Hudson's picture

James:

• The inability to change my zoom once I start using it.

Is that a Mac FL limitation, I wonder? No problem with this in the Windows version.

• The nature of the tool is the alter the entire letter unless I lock down a bunch of points.

Well, yes, that's what it's for: interpolating the position of untouched nodes. The interpolate tool has two states for nodes: touched and untouched. A touched node is one that has been clicked on while the tool is active, whether or not it has also been moved. Touched nodes are, thereform, either moved or locked in their original positions. And all the untouched nodes move relative to the locked nodes, i.e. the position of the untouched nodes is interpolated.

• I have to repeatedly drag points to different potential locations until I get a result I like.

I tend to use the ctrl+click dialogue interface, rather than dragging, but in either case I don't understand why this constitutes a ‘limitation’. You look at your glyph, and you figure out what parts you want moved and in what relationship to other parts, and you touch only the nodes that you need to touch to achieve that result, letting other node positions be interpolated.

There are a couple of things that I do think are genuine limitations of this tool:

1. It isn't dynamically interactive with Fine Preview. I wish it were, so that I could get a nice preview of the results while in the middle of interpolate operations.

2. Moving a whole section of a glyph, e.g. a stem, a uniform distance relative to another part of the glyph always requires touching at least two points. Christian's script seeks to address this limitation, by making it possible to select and shift multiple nodes, but it seems to me that it does it in a way that is overall much less flexible and powerful than the interpolate tool. Consider:

Let's say I have an uppercase U glyph, and I decide I want to make it a little wider by moving the two vertical stems further apart. Using Christian's tool, I can select the nodes making up the left vertical stem, and use the nudge left tool to shift them, say 20 units. But only the positions of the control handles between the bottom selected moved nodes and the nodes defining the bottom of the horizontal curve are interpolated. I have to then go and perform the opposite select and nudge operations on the right side vertical. Using the interpolate tool, all I have to do is lock two nodes on the left vertical and shift two nodes on the right vertical 40 units, and the position of the nodes on the curve is interpolated along with the control handles. That seems way more efficient and useful to me.

k.l.'s picture

John -- ctrl+click dialogue interface

Unfortunately there is a bug in the Mac version of FLS5. Values entered in the dialog are not applied properly. Outlines get distorted. So it is not possible to adjusted outlines in controlled manner which renders this function useless for Mac users. Reported long ago but no update yet.

blank's picture

I don’t think that I’ve done a good job explaining what I actually use nudge for: moving little things. For example, I just used it to drag the tail of a into a better position when creating æ. With nudge I do not have to spend so much time tweaking the shape after the move. It saves me second here and there, and over the course of building out an entire font (and then a family) those seconds start to add up.

Frode Bo Helland's picture

I'm really trying to figure out how to go from regular to bold when the letterform have diagonal curve endings.

hrant's picture

I still don't get what "diagonal curve ending" means.

hhp

Frode Bo Helland's picture

Once again, my type terminology fails me. Forgive me. I’m of course talking about terminals.


(Notice the diagonal terminals!)

Using the interpolation tool to add weight to an "o", or an "n", is no problem. The places I run in to trouble is were diagonals meet each other or a stem (like in "M"), were terminals are cut at an angle (like in my illustration) and usually in tight spaces (like "s", which I have found easier to just completely redraw each time).

.00's picture

Why not just draw it the way you want it to look?

blank's picture

Are you leaving the overlaps in your diagonals so that individual sections of paths can be manipulated without altering adjacent sections?

Frode Bo Helland's picture

James Montalbano: That I have to do in the end anyway, but it would be nice to get an impression of how these terminals would look/behave if I add the same amount of weight all over.

James Puckett: That advice belongs in the “How to” section. I think I just figured out that method on my own (minutes after posting). Do you usually add ink traps at the last stage?

eliason's picture

Wow, that's a great tip, James P!

.00's picture

but it would be nice to get an impression of how these terminals would look/behave if I add the same amount of weight all over.

Why on earth would you want to add the same amount of weight all over?

Perhaps you are after the dreaded "letters dipped in chocolate" syndrome we are all advised to avoid. But why?

blank's picture

Do you usually add ink traps at the last stage?

I try to avoid drawing traditional ink traps because designers can’t be bothered to convert large type to outlines and remove visible traps. Hiding it in a long taper works just as well except in particularly adverse conditions when a specialized type should be used instead. When I do have to draw them in they’re pretty big so I just work them into the letter.

hrant's picture

Maybe not "traditional ones" (I'm not even sure what that means)
but trapping is good for you, and a designer who can't be bothered
to remove them: shouldn't be using a text face so large; and won't
mind the artefacts anyway! :-/

hhp

Frode Bo Helland's picture

James (the terminal one): I don't think I have ever added the same amount of weight all over a glyph, but I often add the same number of units on all verticals (and likewise on horizontals). Is this a case of being reluctant to share your trade secrets? If so, you are under no obligation too and there is no need for an agressive tone. I'm just asking for advice.

blank's picture

Have you tried using KLTF Glyph tweaker to create heavier weights? It’s not perfect, but if you just let it do the heavy lifting and then clean things up yourself it’s a great time-saver.

.00's picture

If so, you are under no obligation too and there is no need for an agressive tone.

I'm sorry you took my comments as agressive. I think we must live in very different places.

Frode Bo Helland's picture

In that case, please forgive me James.

William Berkson's picture

There are in my mind two different issues here. One is time-saving ways to work. And here various software can help to get toward what you want, and what to use is a matter of personal preference.

The other issue is what a good related heavier weight looks like. Here you have a lot of decisions that can't be made well by any formula, which I think it James M's point.

For a start, the visual impact of horizontal and vertical increases in weight will be different. This can be taken into account in software, but all diagonal strokes and arcs will get double doses if you mechanically bold in both horizontal directions, resulting in a lot of ugliness, including clotted joins. I suspect that is the issue Frode was running into, prompting the post.

How to handle the joins is an aesthetic decision that isn't automatic, and this will affect all the arcs as well. How much do you want to preserve relative weights, and how much the same arcs? You can't do both.

A further complication is when you have more than two strokes involved, like a e W M. Then bolding is going to decrease the size of the counters more than in those with two strokes. So again you have to make a choice: do you want to widen the glyph? Widen some of the strokes less than in other characters?

There are other decisions as well, but these are enough to show that there isn't one "right" way to get a bolder weight, and the designer needs to, well, design, as James M. says.

Another issue is how much you want the different weights to look the same. Frutiger pioneered a lot of weights and widths looking similar, with his Univers. But this modernist idea is not the only way to go. Jonathan Hoefler, in his blurb for Knockout, argues that having different looks in different sizes, as type traditionally did, can be a good thing.

Syndicate content Syndicate content