itt ligature - thoughts?

Sye's picture

i was looking at this sample of Gotham:

http://www.typography.com/fonts/font_features.php?featureID=10&productLi...

and thought what about an itt lig. does anybody know if this is ever done?

Si_Daniels's picture

Mitt Romney used Gotham too, right? They were clearly expecting him to be president, not Obama.

Sye's picture

ha!

i've never seen an itt lig, has anybody done it?

Theunis de Jong's picture

I see ft, ftt and it, but no itt.
The 'www' is great! Or do I misread the 'vvvvvvv' ligature?

Nick Cooke's picture

I see the point of the tt lig, but I doubt that's an itt lig.

Here's one I've done:

Nick Cooke

Stephen Rapp's picture

At work I created i_t_i and t_i_t ligatures in a handwriting font, but only to resolve conflicts with crossbars. I don't see a situation where using it in a text face would enhance the design.

Sye's picture

thanks all.

@sii - i can't see the itt...

@nick - thanks, very nice!

@Stephen - i see your point, andyour right, it's not really needed, but is some contexts, like a logo maybe, it could be nice... i'd even be interested to see it helps in reading longer texts, or gets in the way.

James Arboghast's picture

The anser* depends on the font. For a text font an itt lig is of questionable value. In a display font, depending on the design and proportions, they can be of practical and aesthetic value, as can numerous other similar ligature combinations. I put quite a collection of such ligatures into my Rhodaelian font:

As you can see, they serve 2 purposes; they save horizontal space and they look good. Is there any practical advantage to readability or legibility even in a font like Rhodie? Not really, I don't think. I made them because I wanted to.

* I mean, why do we persist in spelling this word "two" ? For that matter why spell "peeple" - "people" ? Makes no sense at all except to those who love hanging onto baggage.

j a m e s

Sye's picture

cool, thanks james. nice. i like the fw.

dberlow's picture

"As you can see, they serve 2 purposes; they save horizontal space and they look good."
The first is obvious, the second, is not. All ligatures are somewhat detrimental to spacing flexibility. In James' example, all of the ligatures are tighter than the solo versions combined. So, were the originals too loose? or are the ligatures, some nearly 50% of normal spacing, too tight? And then there is the "where does it end" question, where the ligatures shown above could only be 1/3 to 1/4 of the ligatures if consistent application in composition is required.

Cheers!

Ale Paul's picture

This is from Feel Script

Basically done to work with contextual.

Best, A.

Sye's picture

i own a licence for feel script! cool!

Si_Daniels's picture

>@sii - i can’t see the itt...

Sorry it's mentioned but not shown... "Additional ligatures in Medium Italic: TT cy itt ry sty ty and tty"

James Arboghast's picture

@Berlow: The first is obvious, the second, is not.

That depends on the point of view of the observer and how flexible the observer is. You're disputing what is aesthetically pleasing. What looks good to the individual is a subjective quality. So your statement is a personal observation, and my assertion that these ligs look good is a personal observation.

Anything anyone cannot handle after that—you are on your own.

No black & white. The solution to a lot of the pointless disputes occurring at Typophile is to recognize that we don't all see the world the same way and just agree to disagree.

That's an arseload easier than getting involved in petty pissing matches.

All ligatures are somewhat detrimental to spacing flexibility. In James’ example, all of the ligatures are tighter than the solo versions combined. So, were the originals too loose? or are the ligatures, some nearly 50% of normal spacing, too tight?

Neither. You misunderstand the nature of the font and lack knowledge of its modes of operation. And if I hadn't shown the unligated letter combos above the ligatures you would not have known they differed. By "were the originals too loose?" you mean to say "are the stand-alone glyphs too loose?" Rhodaelian has several ligature modes. The raw LIGA and DLIG modes apply ligatures with no spacing adjustment of stand-alone glyphs. HIST offers spacing compensation for all stand-alone glyphs to match the density of the ligatures. CALT offers spacing of stand-alone glyphs halfway between normal (uncompensated) and the full compression of HIST mode.

And then there is the “where does it end” question, where the ligatures shown above could only be 1/3 to 1/4 of the ligatures if consistent application in composition is required.

Rhodaelian has more ligatures than those shown. 67 in total. I posted a sample of the units relevant to the thread topic.

j a m e s

Nick Shinn's picture

I haven't created any ligatures per se for this letter combination, but have addressed it through contextuality in several typefaces. The first must be manually applied, the others occur automatically with the default OpenType feature "contextual alternates". Not technically ligatures, but similar in intent.

1. Oneleigh. This is just an alternate t glyph designed to "ligate" with other characters.
2. Duffy Script. Pseudo-random. There are four alternates of each character; for "t", some of them are designed to form ligatures.
3. Handsome Pro. Alternate forms for contextuality, based on the "cancellescera" aesthetic.
4. Softmachine. t with shortened crossbar to preceed f, t, v, w, and y.

Stephen Rapp's picture

"I don’t see a situation where using it in a text face would enhance the design."

Ok, Nick. Your Soft Machine i_t_t ligature is a great solution even if it is subtle. Nice work!

James Arboghast's picture

@Nick: Not technically ligatures, but similar in intent.

I call those combos fitted combos or fitted pairs. I have a question about your Oneleigh itt example—we're looking at an alt t glyph with the swash / ligature coming off the top of it that lines up with (in this example) i and t? And the i in that sequence is an alt dotless i? All in all that's a neat solution.

j a m e s

James Arboghast's picture

@Arbo: HIST offers spacing compensation for all stand-alone glyphs to match the density of the ligatures. CALT offers spacing of stand-alone glyphs halfway between normal (uncompensated) and the full compression of HIST mode.

Actually that was a big fat lie. I made it up to satiate David's conclusion the ligatures in Rhodaelian are spaced at about 50% of the mean spacing value of the stand-alone glyphs. I wanted to see if anyone would realize what a poorly designed font Rhodie would be if it had ligatures that didn't match the density of normal spacing of the stand-alone glyphs and relied on a spacing program to compensate everything else when the ligs are in use. David based his estimate of the density of the ligatures and arrived at a normal spacing value based on a selective picture showing ligatures only. When you can see more the rest of the font it makes a lot more sense and you'd be less likely to jump to an incorrect conclusion.

As you can see in this sample, it isn't like that at all. I designed the ligs in Rhodie to match the normal spacing and stroke density. No spacing compensation needed for those ligatures. No ma'am.

Nyah, just windin' you up ;^)

j a m e s

Nick Shinn's picture

...question about your Oneleigh itt example...

The "swash t" was originally designed as a step towards making the classic ct and st ligatures.
Then I realized it would also work with other characters (with some judicious kerning).
The "itt" is not really one of the better examples; "kt", "rt" and "xt" work quite well, though.
(Yes, it is dotless i--an old trick from the Diatronic days, where "f-dotless i" was used by type houses to make a ligature of sorts.)

John Hudson's picture

David: All ligatures are somewhat detrimental to spacing flexibility.

But not everything that looks like a ligature is a ligature, and some apparent ligatures space quite well within reasonable constraints:

[+20 tracking is getting beyond the reasonable limit in my opinion; and in the case of this font the ligature substitution would normally turn off beyond +16 units (see below).]

I've given up on ligatures in the traditional typographic sense of a single sort (glyph, piece of metal, etc.) representing more than one letter, and have moved to using ligating contextual alternates.

This mechanism also makes it possible to cater for ligation of arbitrary diacritic letter sequences, which is useful if, like me, one is making fonts for more than European languages and can't always anticipate what users will need or want.

By putting the contextual lookups for these substitutions withing the layout feature, one can take advantage of InDesign's tracking tolerances for ligatures (and one can also design the ligating glyphs with those in mind. At TypeCon this year, I asked Eric Menninga about these tolerances, i.e. the point at which InDesign disables the feature for tracked text. He checked, and reported that they are not fixed tracking units but are based on percentages of the width of the word space glyph of the particular font: +8% or -22%. So if a font has a word space width of 200 units then the feature will switch of beyond +16 units tracking and -44 units tracking.

James Arboghast's picture

The “itt” is not really one of the better examples; “kt”, “rt” and “xt” work quite well, though.

Okay, I will have a look at those other ligs. The interesting thing about the Oneleigh "swash t" combinations is how far the font can be tracked out or in before the alignment starts to run out of usable range. No matter which solution a type designer chooses to use—a fixed width ligature or a kerned / fitted pair—tracking the font in or out a sufficient amount can still compromize the type designer's ideal range for stroke density versus whitespace, potentially messing up color.

j a m e s

dberlow's picture

Any thoughts? you asked.

James: "No black & white." Really? Another way of putting it is that when the internal and external spaces are in a state of conflict for superiority, no one should have to compose it, not minding the ligatures or the reader. Before deciding if that's subjective or objective, don't you think you should look?

I agree that there are a lot of pointless disputes here. But there are also quite a few disputable design decisions as well. Not this design, I like it in theory, but what's the point behind the variety in the upper right corner of n and m?

"David based his estimate [...] on a selective picture showing ligatures only." Not precisely, I based it on the difference between the ligated and unligated versions of the same pairs, you kindly showed.

John: I like your solution, but if you can track ligatures so they are not a problem, then they are not actually ligatures are they? :)

Cheers!

Nick Shinn's picture

Nice work, John.

This solution addresses David's concern about tracking messing up spacing flexibility.
For instance, suppose one considers that the spacing of vertical strokes between f and i should be slightly greater than between l and l, then "ligating alternates" preserve this relationship through tighter and looser tracking:

So if a font has a word space width of 200 units then the feature will switch of beyond +16 units tracking and -44 units tracking.

Thanks, that's good to know.
This switch-off is useful for ligatures, but not for pseudo-randomization.

John Hudson's picture

Better Than Ligatures™

I Can't Believe It's Not A Ligature™

dberlow's picture

Nick: This solution addresses David’s concern about tracking messing up spacing

It does? I think this is a great solution where there are serifs, and lots and lots of white space to the black space. However, I think both you and John are showing solutions to a problem that does not exist in the origins of this thread.

Cheers!

Christian Robertson's picture

I've always been a fan of the tt/TT lig. They feel like they should be stuck together.

ebensorkin's picture

I like the CALT for Ligatures model John describes in his post because it gives you flexibility in layout and far more control of the white spaces/shapes.

To go back to the gotham example ( is that what you mean David?) , if the t shapes and side bearings were different at the end of a word vs when they were inside a word, you could improve the spacing. Or better still, if the shape of the t depending on the specific shapea next to it you could control the relationships/spacing/white shapes even more precisely. And if that were the case, then, from my point of view if the solution happens to be a ligature or not is suddenly a style question not a spacing question which is a far more ideal state of affairs.

Syndicate content Syndicate content