w_r ligature in script font

Grzegorz Luk (gluk)

I have doubts about "w_r" ligature in my script font. On image below example in word "wrong".

It's readable or reminds too much "ur"?

Christian Thalmann

Looks legible to me... but try setting it (and /u/r/) in a few nonsense words to see whether it's just intuition from context.

Grzegorz Luk (gluk)

Christian Thalmann thank you, some more words..

and nonsense word

Jasper de Waard

The nonsense word is difficult, but everything else is easy to read. aurewrty?

Grzegorz Luk (gluk)

Jasper de Waard  aurewrty?

yes, thank you.

maybe in non english words too?

Christian Thalmann

Admittedly, I had a bit of trouble reading aurewrty at first sight. I'm not sure whether the unexpected serifs in /n/ and /u/ might interfere as well.

Didn't have the patience to wait for the animated sample. The Polish is certainly more difficult for me than it would be for a Polish person.

Joe Elwell

I don’t think the problem is in the ligature itself but, the construction of the diagonal characters, altogether. They are quite homogeneous with the rounding at the bottom, which is fine, but not very distinguishable considering the similarity of the top halfs of the /u /w and y. (What does the the /v look like?)

The top horizontal of the /z and looping of other characters are interesting, as are the serifing of the /c and /s. Should the /u more closely the /n? Thus, I am wondering how maybe some of those features could be applied to your diagonal characters which will help it read clearer in those tough ligature situations. Try test strings of uuvuu vvuvv uuwuu wwuww wwvww vvwvv and see how quickly you, or someone else, can spot the outlying character. If you can’t, try a different construction.

Designing a thin, mono-linear, script face to be legible is a nice challenge. Good luck!

André Simard

I also have to admitted that it's really hard to read. Keep going your test.

Grzegorz Luk (gluk)

@Joe Elwell  @André Simard  @Christian Thalmann  Thank You

Joe Elwell said:

... Try test strings of uuvuu vvuvv uuwuu wwuww wwvww vvwvv ...

unfortunately in examples above glyph /w occurred  only as contexual alternate. Base glyphs /w /v look like bellow

Joe Elwell

I think you need to work on the /u first. (Which may mean you need to go back to work on the non-contextual /n, as well — It looks too wide at the end of the word can) That seems to be the weak point in these strings. Use the nnunn oouoo strings to test rhythm and texture. I like the balance of the v-w strings and think those should serve as a reference point for the sort of rhythm you want to achieve across the board.

Grzegorz Luk (gluk)

Joe Elwell said: ...It looks too wide at the end of the word can) That seems to be the weak point in these strings. Use the nnunn oouoo strings to test rhythm and texture.

/n at the end of the word "can" is is contextual alternate (end form, like last /t in "that" ), and is intentionally wider as standard /n

Joe Elwell

In that case the contextual (final form) /n is too wide. You can treat the exit stroke with more of a swash ending if you like but, the counter shouldn’t be wider. 

Thomas Phinney

My brain can’t get over the added horizontal serifs that break the connecting-stroke rhythm. /w is fine as the first letter in a word, but not when it is in the middle of the word. Same for /n.

Do I just need to get over it, and used to the idea that it is a hybrid of a printed font and a handwriting style font?

Vasil Stanev

Either add weight to some parts of the letters or look how it's done in Bukhari by Microjihad Inc. and Nickainley by Seniours Studio (both of which I did the Cyrillics for). There is also my Bronn Rust, somewhat less developed on my part.

Grzegorz Luk (gluk)

@Thomas Phinney hybrid is most probably best description for this. None of the glyhps was handwritten, but geometrically constructed. My Idea was geometrical, good readable typeface with handwriting feel... horizontal serifs are consequence of "readability"

@Vasil Stanev  thank You, but this font stay thin & mono-linear

@Joe Elwell the longer I look at "can", I think You are right. Terminal /n is too wide. Most probably I make terminal /n only a bit wider as standard /n. 

John Hudson

I think you've got a ductus problem, in that you are merging the final stroke of the w and the initial stroke of the r into a single element, so the result is readable as ur or wı. 

Adam Jagosz

Poles don't know this form of /r from school, so it is not surprising that we take the loop for just an instroke. In reality it is integral to the letter, Romanians might even write /r like this:

Personally I even think this calls for locl, as this form is totally alien for Polish use. (Unless you're going for foreign, or chic, vibe).

Adam Jagosz

On the other hand, merging /r with /z is (I think) fine for Polish. Not sure about other languages.

Joe Elwell

I think below is sort of what @John Hudson’s point is about the w_r ductus... the loop belongs to the w as a natural transition to the next character. (This is an italic style I just started transitioning into a modern script-face)


And I completely agree with @Thomas Phinney about losing the serif on the straight stems. It isn’t natural within handwriting — I think you are mistaking readability for legibility in this case; and geometrical as a hand-written script style is an interesting concept. Perhaps focus on what is most important for this design to evoke and incorporate the other aspect where it make sense / works best. 

Below is quick scrawl of how your /n (sans serif) and your loop from the /z & r could inform others like the /v and /w and other, as well.


If anyone thinks I am wrong, please say, I won’t be offended.

John Hudson

Just noticed that the dotless i in my last post isn't displaying. I meant that the wr combination in the original post can be read as ur or wi (without the dot on the i), because of the way the stroke patterns of the two letters are merged.

Adam Jagosz

Yes, can we get a better version of Alright Sans? This one has empty glyphs for dotless i, pi, and probably a whole lot of other glyphs, plus it has no Vietnamese, no Greek and Cyrillic, not to mention Armenian. (Maybe there's another typeface that would suit TD better? For example... a system font?)

Nick Cooke

Thomas Phinney

That’s a lovely saying, but it seems self-contradictory.
1) Anything goes IF it does (a) and (b).
2) Nothing is considered “wrong.”

Nick Cooke

looking at it now I can see it needs another th lig, an re and another g.end. 

Adam Jagosz

Some more test words: awry awrah chowry cowrie dowry ewre heathwren inwrite Jewry Lawrie Lowrie miswrite wranny wrasse.

You asked whether it's readable: it is. I don't think most readers will care, but... there's something intrinsically wrong about it, kind of as though you'd start writing Cyrillic ell or em connected directly to the preceding letter, without making the little upstroke/loop:

(And you get an i instead of em.)

If you want to go through with this, it would be best justified if it fixed a problem. I'd love it then if you showed us the problem!

Adam Jagosz

Regarding legibility, you might consider tapering the outstrokes a bit (or a bit more) while leaving the main letterform monolineal. This might eradicate the need for the serifs.

(Unless this is strictly a skeletal font; what are you animating it with?)

Nick Cooke

Thomas Phinney said:

That’s a lovely saying, but it seems self-contradictory.
1) Anything goes IF it does (a) and (b).
2) Nothing is considered “wrong.”

Maybe I should have said ‘when’, not ‘if’.

Grzegorz Luk (gluk)

Adam Jagosz said:

... there's something intrinsically wrong about it...

most probably there are more things in this font, e.g. ligatures S_h, T_i - who writes like that...

Adam Jagosz said:

... you might consider tapering the outstrokes a bit (or a bit more)...

I create this font for SVG animations (with on-line images generator), that mean it  stay thin & mono-linear. But legibility (thx @Joe Elwell ) is important.

@Nick Cooke: I like Your Idea, btw. script font (Goskar ?) too.

Nick Cooke

Grzegorz, your latest wr sample looks like wi (with idotless). You need a smallish end loop on the w and a bigger initial loop on the r. Thanks, yes it is Goskar which is progressing not exactly speedily, but it’s getting there. This is the lightest of 4 weights. 

Grzegorz Luk (gluk)

@Nick Cooke  yes, mistake with w_dotlessi is possible. thank you,(@John Hudson  too)

First try of new w_r ligature:

Nick Cooke

It’s better, but looks disconnected. Move the loop away from the w, closer to the r.