T.T.P.K.A. Omnium

This is The-typeface-potentially-known-as-Omnium.

It is a text typeface with a rotated humanist axis, designed with a large x-height with the aim of strong horizontal emphasis and even colour. A sanserif sibling is also under development. It is my first typeface.

I’d love to hear your impressions, suggestions, criticism.

I became acquainted with Fontlab for the first time last week so my spacing leaves a little to be desired. Let me know if you’d like to see samples in any other format.




Dan Milne
13 Aug 2005 — 4:42am

I like it. Here are some suggestions:

I'd put more curve into the top of the s.
The bottom right of the b feels uncomfortable.
The terminal of the beak of the a (top) could be rethought, as could be the bottom of the g's descender.
The r's terminal could be heavier.
The v's crotch should be darker.

Hope that helps!

Hi Dan.

Many thanks for your suggestions. I had a feeling I should soften that curve on the b. What do you suggest for the terminals on the a and g? An s-style serif terminal perhaps? Or is it more a matter of breaking the stroke at a different angle?

My PDF file contains the full alphabet and ranging numerals. I'm not sure why it didn't work. It can be viewed at:



first of all, i should ask: what is the purpose of this typeface? any particular applications? This would effect what advice i would give in making revisions. For my comments i'm assuming it's a standard face for text.

wow! i didn't even read dan's comments till after i looked at the face and his comments are the same as what mine would have been. just a couple more things:

The bottom right of the b feels uncomfortable.
same goes for the top left of the bowl of the g (as i'm suspected it's the same piece, only rotated)
the inner counter of the o seems a bit too boxy. maybe smooth out the curves at 1 and 7 o'clock.
the bowl on the a could be a bit more angular/contrasty to match the other curved elements. and to solve the a's terminal problem, you might just make it a bit heavier on the end.
is that m a doubled n? even if not, it seems a bit too wide still.
i think that poor little s feels very uncomfortable. try to make it less disjointed.
looking at the pdf:
all the terminals look a bit hesitant (they go out with a whimper instead of a bang) i'm talking about a,f,g,j & y.
the w has the same crotch problem as the v, as does the y.
The caps are very nice. Terminal problem on J. The stroke on the Q looks like it's not sure it should be there. Are those ink traps on the A, M, N, V, W (and v, w, y) if so i think they need to be re-thought so as to give more weight to the crotches (as dan stated before). The C seems a bit squonched, but i'm not sure exactly to suggest to fix that. The S and the Z seem to be leaning to the left. There should be more difference between the bowls on the P and the R.
My preference is for slightly longer descenders, but that could be just my personal preference.
The figures seem a bit small. i think taking the height up between the x-height and the cap height they'd look better. If you decided to make the descenders longer this would help the numerals, IMO.

And as with all critiques by yours truly, a disclaimer: I am no expert, all advice is my own personal opinions and does not reflect the views of typophile and is to be taken with a healthy dose of salt. use this information at your own risk. i refuse to be liable for consequences caused by the adherence to suggestions or the lack thereof.

it looks like a huge success for a first typeface! where did you learn typeface design an how did you work on that one? did you first sketch the letters and then scan them?


Do you try to establish broken forms into a common roman typeface? I think yes and I like it. So, please don't round up all the forms. To judge it, we have to see it on real text in small point sizes.

Some typesfaces looking very strong if you look very close to it but working well as bodytypes.

ps. I can't download the pdf.

-- astype.de--

i didn't mean to round everything up, but some curves do need a bit of smoothing out, IMO.

Hi Paul,

Thankyou very much for taking the time to look at my typeface and for your detailed suggestions. They are very helpful. Apologies for my delayed reply: it is early Sunday morning here in Melbourne.

Omnium is intended for reasonably extended setting in magazines and books. I think it has probably ended up with too much personality for serious continuous setting.

I have been steadily rounding the counters from what started out as a much more extreme typeface. I’d like to remove some of that boxy discomfort without losing too much of the angularity in the counters which gives the typeface much of its personality. As you suggest, only some of the counters are problematic.

I’ll give some more presence to the terminals and the crotches and deal with soften the more awkward curves. I prized the crotches apart (so to speak) in order to achieve even colour -- probably at the expense of the strength in some of the intersections. I suppose I can keep the ink traps to some extent. They just need to be more subtle.

You guessed correctly on the doubled-n m. A trap for young players, I guess.

My intention with the old style numerals is to find a size which works well within text with lowercase but which also works with longer strings of numerals in their own right. I find that the height of old style numerals often jumps about too much with certain combinations like 345 688 120 (see http://omnium.danmilne.fastmail.fm/files/num_demo.jpg ).

When you suggest taking the height of the numerals up, do you suggest increasing all numerals or making the ascenders ascend more and the descenders descend more? I guess I’ll try some options and see what works. I also intend to develop a cap-height lining set.

Hi Oribendor,

Thanks very much for your words of encouragement. The type design scene here in Melbourne is small but I’ve had plenty of advice from local typographers and type designers with more experience than myself. I have spent quite a lot of time looking at Dutch typefaces and I think a lot can be learned from sites like this one!

As you guessed, I sketched the letters first and as Andreas suggested, the curves are made up of broken forms which I am trying to reconcile into flowing curves. I’ve attached a sample of the kinds of drawings upon which my letterforms are based: http://omnium.danmilne.fastmail.fm/files/omnium_drawing.jpg

Hi Andreas,

Sorry you haven’t been to access the PDF. Hopefully it can be viewed at: http://omnium.danmilne.fastmail.fm/files/Omnium_serif_28_metrics.pdf

Once again, thankyou for your contributions.

I'll make some adjustments to the typeface and post an updated version in the next couple of days.


numerals: yeah, maybe you just need to make your descending characters descend all the way. the 3,5,7,9 all seem a bit tiny
and yeah, make the ascenders on 6 and 8 rise to the cap height.

Hey, you're using my "metagram"! Cool. In case somebody (more like virtually everybody) doesn't know, a metagram is a fancy pangram that lets you evaluate a font much better:

Anyway, onto this quite interesting font...

First of all, the PDF is much more convincing than that GIF up top, so I'm glad you've made that available. This face has quite enviable poise. And I don't agree that "it has probably ended up with too much personality" to serve as a text face; as Schwartz's Houston Chronicle effort has hopefully proven, the text face envelope still has huge room for expansion, although in a direction seldom explored. That said, I think this still needs a lot of work to become adequate.

My foremost "macro" critique is that the color is a bit light for the vertical proportions, which seem ideal for around 8 point setting. If that's too small a point size in terms of your target, I'd increase the ascenders a bit, shooting for 9 point; but I might still increase the color slightly. If you like shooting for 8 point, I would pull in the descenders a hair.

Micro stuff:
"a": It needs to be happier - it's a very important character. I don't think the terminal of the top is the problem (it's certainly better than the one in the "g") but probably that the top of the bowl shouldn't be so horizontal - try tilting it gently.
"f": Could use a stronger beak.
"g": Give its terminal a serif, like in the "c". Or much better: make a binocular one... with an open bottom. But if you keep it mono, give it a nice rising ear.
"j": That chirographic terminal seems to actually work here.
"k": Best glyph so far!
"p": Bottom of the bowl too heavy.
"q": depending on what you do with the "g", consider a pointy ear.
"r": Majorly needs a powerful beak.
"s": At first I thought it was too "coiled up", but now I think it's da bomb.
"v", "w": Too much separation at the diagonal joins.
"y": As in the "j", the terminal works here.
"B": Slightly too wide.
"C": A bit too narrow.
"E": A bit too wide.
"F": I'd lower the bar.
"G": Slightly too narrow.
"J": Tail not working at all. Maybe descend it (if you end up descending the "Q").
"K": Nice, but a bit wide.
"M": Too wide.
"P": Slightly larger head.
"Q": Tail not working, especially inside. Not descend it?
"U": A bit too narrow.
"V", "W": Same problem as in the lc.

I can only recommend what I so often do for text faces: hybrid numerals! Actually not far from what you already have, but taller than the lc x-height in body.

All in all, great stuff - please keep it up!


Hi Hrant.

Thankyou for your detailed critique and words of encouragement: lots of things for me to think about.

In terms of proportion and colour I think I have been too vague in stating my goals – sorry. Rather than 8 or 9pt regular, this typeface should probably be seen as a semi-light for setting at 10pt. I intend to create a regular, semi bold and bold in the future. I started with the lightest weight because I’m told that this is the most difficult to establish in terms of curves and stroke variation.

Therefore, as at 10pt semi-light, I imagine I should increase my ascenders somewhat and maintain the same level of colour? Further down the track, I’d like to create a caption cut with reduced extenders for setting at smaller sizes.

The colour of Omnium has simmered down during the testing process. I’m aiming to find the smallest amount of stroke variation which still conveys its rotated humanist axis at 10pt (with laser output.)

With regards to proportions, a small preliminary comparative analysis I did may be of interest:

Before I started drawing Omnium, I made a rough analysis of a bunch of typefaces in terms of proportion and colour, making measurements and comparing them as percentages of each typeface’s x-height (which I believe to be the most useful benchmark in terms of comparison). I measured ascender and descender lengths, maximum stroke thickness (typically the stem of the lc d), minimum stroke thickness (crossbar on the lc e). I divided min stroke thickness by max stroke thickness to give me a crude measurement for comparing the variation stroke thickness. To measure how condensed each typeface was, I measured the widths of lc d, e and o. I also measured the axis of stroke variation for each. A table of data can be viewed here:


Yes. I am very keen on the idea of hybrid numerals: not to replace the lining numerals, but to replace the old style ones. I’ll explore some possibilities.

Thanks again for your advice. I’ll post a revised version ASAP.


For 10 point, I'd definitely lengthen the ascenders a bit. But this sort of thing isn't done to maintain color, it's done to be in tune with immersive reading, which is helped by different proportions depending on size.

> I started with the lightest weight because I’m told that this is the
> most difficult to establish in terms of curves and stroke variation.

Maybe. But I'd rather see the design of the Regular done first, simply because that'll be used most. Note that basing things on an extreme weight can skew the Regular in peculiar ways: potentially interesting for display, generally bad for text. On the other hand, you're not really far from what I personally would like to see as the Regular weight of this design (a bit darker) so I don't think you'll have any such problems.

> http://danmilne.fastmail.fm/omnium_decon.pdf



I have made some changes to Omnium U&lc and numerals. A PDF can be viewed here:



I have lengthened the ascenders by 3%, the descenders by 1%. The Cap-height and stroke weights are the same as before.

I tilted the bowl of the lc a down and gave it a serif terminal.

I have given g, j, f, r, y more weight and more ‘scoop’ on their terminals. I am hoping to avoid a serif terminal on the tail of the lc g.

I rounded the lc o ever so slightly.

I narrowed the lc m.

I put small rising ears on lc q and g, removing the unnatural serif from the g (much better for letterfit now)

I put a ‘downward ear’ on the base of the lc b and opened the curve, Hopefully the awkwardness of the lower right of the lc b is now vanquished.

v, w, y, M, N and W now have more weight in their crotches.

I have squared up the corners of ‘s’ in U&lc.

Minor adjustments to the proportions of BCEFGKMPRU

Added a serif terminal to J. I am keen to keep Q and J above the line – J at the very least.


J is now wider, more stable. What thoughts?

The new Q needs work. (The current form was inspired by Hrant’s prototypes on the Patria thread – thankyou). Too much like a flat tyre at the moment? Given the conformity of the other caps, I’m inclined to head towards a simple descending Q.

Lowercase numerals:
I‘ve opted for lining 2/3 height numerals. I experimented with ranging numerals from that 2/3 height but numbers like 3 and 5 were floating on either side of the x-height / baseline so I ditched them for the lining option. I intend to make my small caps this size too. What thoughts?

Does my lc t ascend enough?

Any other observations?

Many thanks,



The uppercase Q is both unpleasant and too close to an O. If it had a long tail that went below the line, it would be much better, IMHO.

As for the numerlas, I don't think it works. They should be either bigger than lowercase or old style (that is, have ascenders and descenders, like in Georgia for example).

Other than that, there may be some small corrections still needed to be done, but the overall look is quiet impressive!


I think it's really starting to sing now!

The "c" is narrow.

I would give the "f" still a lot more beak. Your x-height is pretty big, don't worry about touching.

Numerals: "3/4 lining" ones are my second favorite. And seeing them in action I'm thinking they suit the certain "business-like" mien of this design quite well. BTW, I would give the diagonal of the "7" a bit more character.

Caps: The "A", "E" and "S" are a bit wide. The "J" looks great. If you're sure about restricting the caps to above the baseline, I actually like that "Q", although I would try to make the tail longer - and maybe give it a strong flaring terminal, not a point. BTW, if you do end up descending the "Q", you really might as well descend the "J" as well - for one thing that helps spacing, especially for Slovenians. :-)

Spacing: a bit loose overall.


Here’s another update of Omnium Serif.


I’ve added the primary punctuation / symbols. I’ve adjusted a few character widths/terminals, tweaked the Q and tightened the spacing a bit (but probably not enough??).

I've strengthened the beak on the f, but I'd really like to keep it narrow to avoid the need for ligs. (The substituted ligatures in Indesign dont work when small text is tracked slightly).

I will make descending alternates (perhaps even default) for J and Q. It'll make life easier for the kids from LJUBLJANA (Thanks, Hrant). I think the above-the-baseline Q is working better now with a longer tail and more weight in the right places. Hopefully there’ll be no confusing it with the O?

Is the spacing tight enough yet? I think it’s probably still too loose for 10pt? There is still no kerning.

Is the ampersand a bit dark? Do you think I need a serif on my sharp s? Any obvious problems with the analphabetic characters thus far?

Once again, thankyou very much for your contributions. I’ll post a PDF of Omnium’s nerdy new-born sibling over in the Sans Department in the next few days.

Cheers, Dan

I've just posted Omnium Sans in the amputees wing. He's a little more subdued than his big sister but I think they make good playmates.


I'd love to hear what you think.



Some minor details could be improved (like the C, for example - it looks narrow and tall), but overall good work.

About the sans: I think the term "amputee" is quite appropriate... Seriously, the serif already has a "sans" quality to it, especially because of the low contrast between thicks and thins; I think you could focus on improving the serif instead of trying to derivate a sans. IMHO.

Hi Eduardo.

Thanks for your feedback. I am actually quite pleased that you say my serif has a sanserif quality to it. It was always my intention to create a family of hybrids: to push the serif and sanserif designs as close together as possible… to make a serif which takes proportions and geometry from sanserifs without losing too many of the serif features which help readability: the humanist axis, the stroke variation and, of course, the serifs.

In the same way, I’d like to keep as much of this humanist readability in the sanserif as is possible without straying too far from what is considered a classic sanserif. So I’ve been waiting to stabilise the design of the serif before (gleefully) amputating the serifs, reducing stroke variation and making other necessary adjustments.

So I'm not really deriving a sanserif from my serif. I like to think of them as twins who have grown up together :-)

I started with more variation in stroke thickness in the serif, but the boxy, angular counters were looking very awkward and I was forced to retreat.

If my serif appears to be too much of a ‘sanserif with serifs’, how do you suggest I improve it without increasing the stroke variation? I guess I could bracket the serifs a bit more?


These are the small caps for Omnium Serif:


I have made them to match the height of the ¾ lining numerals. Perhaps they are a little larger than is ideal, but I think they will be more useful if they match the height of the numerals than if they sit somewhere in between and I dare not make the numbers any smaller.

What do you think?

I’m having a lot of trouble getting the M, N, A and W (in small caps) light enough to match the other letters. Do you think I should further reduce the weight of the strokes – or is consistency of stroke weight more important than evenness of colour?



> I have made them to match the height of the ¾ lining numerals.

Smart. And I don't think they're too big.

But something else: if you're going to keep this overall loose fit, you'll need to make your blank space wider (and use more leading).


Here’s my first attempt at Omnium Italic. Her proportions are based upon a 90% width and 12-degree sloping of the roman.


She ain’t elegant, but I think she matches the roman quite well. I’d be very grateful for any feedback you can provide.



I like the italics; very simple and functional. Quick comments:
- e: could have a slightly narrower/smaller bowl;
- f: IMO, the long straight descender is disturbing, a curved descender would work better (or maybe no descender at all?);
- x: too wide.

About my previous post (regarding your sans version): it's just my personal opinion that sans-serifs directly derived from serif typefaces are often unnecessary/dull. I think this serif typeface is fine, that it has a certain "sans" quality to it is a good thing. But the sans needs more elaboration.

-- omine.net

It does match the Roman very well. I might make it slightly more "delicate" though - maybe make some counters (like of the "e") smaller, and give some letters (like the "s") smaller heads - and maybe make the "a" the classic WAD form with the narrow top (which Parkinson admires, IIRC).


Here’s a revised version of the italic.


I have reduced the counters and narrowed the heads of many of the characters (Thankyou Eduardo, Hrant). The shapes are a bit softer and more cursive overall. I have also streamlined and simplified terminals on the curved extenders and added more curve to the tail of the ‘f’.

I have made ranging default numerals as well as a lining set to go with the (hopefully seldom-used) small caps. I have also loosened the spacing slightly. Do you think the spacing is ok?

Hi Eduardo,

Thanks for your advice once again. I agree with you that Omnium Sans needs some work. It is clearly derivative of the serif and I think that as an individual typeface it struggles. As a supporting face for the serif it seems quite suitable, but I would like to see it rise above this level. I will turn my hand to it over the next few days.

Hi Hrant,

Firstly, thank you for your consistent, constructive advice. The design of Omnium owes you much! It is amazing how much difference the tweaking of an ‘s’ can make to the personality of a typeface. I explored the idea of the Dwiggins narrow-topped ‘a’, but I couldn’t make it work so I opted for a compromise in-between and adapted all the bowled characters.



I think this is a big improvement.
That new "f" is especially nice.


one quick observation: the ligatures seem too wide.

These are both very beautiful faces, Dan. At first viewing, I thought it would be ideal for a magazine set, and then read that in one of your subsequent posts.
I have art directed several (U.S. circulation) magazines, and I think your design would hold up well in that kind of context; it seems particularly suited to aqueous coated stock.

hi! I would like to say that your shapes are quite interesting. What was your 'fuse'? I can sense some similarities with Frutiger's Apollo or Blockland's Proforma, though it it is distinct and very well optically balanced in both shapes.

Keep the good work.
(what was the software you used?, and have you considered glyphs for alternate characters?)

Hi seml,

Thanks for your kind words.

Silmbach’s ITC Slimbach, Bilak’s Fedra Serif, Unger’s Demos & Praxis, Carl Dair’s Cartier and Bloemsma’s Legato all influenced me in various ways.

Blokland’s Proforma is a great typeface. I was introduced to it after I did my first drawings for Omnium. It gave me confidence to keep the serifs long and unbracketed. I have just looked at it again and I think there might be some ideas there which will help me improve my italic.

I wasn’t aware of Apollo, but I just checked it out and it looks like Frutiger had similar goals with his long baseline serifs and his very oblique axis of stroke variation.

I drew my characters by hand and then in Illustrator before taking them into Fontlab. In retrospect, I should have taken them straight from hand drawings into Fontlab because it is so much more precise than illustrator and much easier to test your letters as you draw them.

So far I’ve only made alternates for ‘Q’ and ‘J’. I originally planned to keep all my caps above the baseline, but I was never really happy with my ‘Q’ and the above-the-line ‘J’ makes spacing difficult so I developed these.

Do you have some suggestions for other alternates?

My f-ligatures are simply joined at the crossbar, so I’ve used context dependent substitutions to substitute a single long crossbar’d ‘f’ with the existing I such that the letter combinations can be tracked freely… As opposed to stiff ligatures. If that makes sense…


Bringhurst has written that we shouldn´t try to make vain changes in what history has proven to work. Changes sometimes only scramble the code that is known for so many... therefore I think that the choice to draw the 'Q' and the 'J' below the baseline is "typographically correct". Although we look for changes in order to distinguish "our" drawings from others, is it also known that only the very careful and microscopic eye can distinguish the novelties. It has to be banal enough for you to read without changing speed and recognition, but somehow leaves a perfume of a personality that gives it taste.

As for alternates, in roman type, the 'c's and the 't's can always fall is love, just like de 's's and 'c's, and the 's's and 't's. My advice is for you to look at the forms Slimbach made in Adobe Garamond, which drinks from Griffo's work in the Renaissance. But what I believe to be more interesting is the personality of the italic letter. There's a wonderful study in the work, also by Slimbach - Poetica. I had the honour of exploring this family on a small book on Old Music, through the combination of alternates, small caps, swash capitals and the chanceries, working aside with Bembo.
We should always learn from old ways. (Mrs Eaves by Licko shows that particular interest).

Your typeface Omnium, in particular seems a challenge for exploring such a feature. I liked very much because it reminds me of the faces I have referred, but it has a touch of the geometric modernism that makes it interesting.

I don´t know how valid you might think my critics are, since I haven´t drawn a typeface so far. I just a guy who loves letters and history and art, and the way that they interact.

As I have said before, keep up the good work ;) and keep me posted. I'd like to see the evolution.

Ligatures can have two roles: style and function. ‘s_t’ and ‘s_c’ ligatures are examples of the former; ‘f_f_l’ is an example of the latter. As a typeface designed for magazines and books, Omnium concerns itself only with functional ligatures which solve spacing problems. The ‘s_t’ ligature is beautiful, but unnecessary and can be seen as a distraction in immersive reading. Omnium’s ‘f_f_l’ ligature is almost entirely at the function end of the scale. It solves the problem of the undesirably close ‘f’ and ‘l’, adjusting the character shapes slightly to allow comfortable spacing. Rather than taking a purely functional approach to the f_f combination, I chose to unite, rather than shorten, the crossbar between the two. Perhaps a shortening of the offending crossbars (hence avoiding any binding of the letters) would have been a more ‘invisible’ solution, however the long crossbar seemed more in tune with my goal of horizontal emphasis. And can a ligature still be called a ligature if it does not bind, but serves only to space?

I hope that, in avoiding distractive ligatures, Omnium is not considered bland.

For similar reasons, I have replaced the original non-descending ‘Q’ with a more standard and predictable design. However, I have kept the original non-descending ‘Q’ as an alternate for the odd occasion when someone might like to set a heading in caps without descenders, perhaps with a rule below.

I am pleased that Omnium reminds you of Garamond and Griffo. In fact, the name Omnium comes from that much referred-to page from Jenson in De Evangelica Preraratione, Venice, 1469. It is a word chosen both for its ‘physical’ beauty and for its meaning: it is the genitive plural of Latin omnis, meaning all or of all. But it is the Jenson text’s rhythmic nature and the strong horizontal lines, rather than the beautiful ligatures, which have contributed to the design of Omnium. As a straight-forward and business-like typeface, Omnium may not benefit from such stylistic ligatures. Perhaps they can be developed as a side-project. As you suggest, it would be a great challenge to design harmonious ligatures for such a rigid, geometric typeface.

>therefore I think that the choice to draw the ‘Q’ and the ‘J’ below the baseline is “typographically correct”

The descending ‘Q’ and ‘J’ should not be seen as more ‘typographically correct’. The descending ‘J’ makes it easier to achieve even spacing (with fewer white space anomalies which disturb the horizontal movement of the reader’s eyes), but I would argue that as a form of the ‘J’ it is no more conventional than the non-descending model. Even as I draft this post in Microsoft Word, I notice that Morison’s ever-present Times New Roman sports a non-descending ‘J’. And most sanserifs as well… Omnium will use non-descending ‘J’s at the start of words and descending ‘J’s otherwise.


> can a ligature still be called a ligature if
> it does not bind, but serves only to space?

Good question. If you consider the white as important as the
black (or really, important at all) then I would say yes, totally.


I' glad to see that you understand the meaning of my words. You would be suprised with the rude knowledge on typography that I have to deal with ;).

In fact, the reason I said that was a challenge is precisely because the 'omnium' face runs to purity of form that Tschichold and Gill seeked, which makes it difficult, or should I say 'inadequate' to treat it like a classical typeface. (Just trying to defend the experiment and the 'errare humanum est'.)

The apprecitions I've made had the goal of taking to the limit the articulation of influences that you are 'printing' with your face.
It gives shape to what I whould draw, that's why I couldn´t look at it without tell you something.

(And I agree that the name is well applied, considering the basis of this kind of drawing).

I whould buy it.
Keep up. ;)

I’ve recut omnium italic.

I have adjusted each letter so that the slope is more consistent (thanks Wendy).

I have made it considerably more cursive. The bowls now form complete arcs, cutting into the stems slightly. I hope the intersections are not too black?

I have also reduced the weight of the serifs because I felt they were disruptive before..

I think it is looking much better, but I am concerned that it no longer matches the structure of the roman as well as the former design.

I’d love to hear your thoughts.

Cheers, Dan

Here is the latest version of the italic.

I have softened many of the character shapes so that they better match the serif and have removed those awkward top right-corners on the a, d, g etc, opting to expose the top of the stem instead (Thanks Brad). I have adjusted curves on the ‘m’ and ‘n’ as well.

Only two things I'd like to comment: the finishings of the 's' look a bit strong when compared with the finishing strokes of other letters e.g. the leg of the 'k', the one down right. The serif detail looks different. Is that on purpose or are you afraid of losing the squared shape? Also, that same purpose seems to affect the bowls like the one of the 'a', for instance. It gives it a broken shape, as intended, but I believe that the horizontal elongation should be enough to give the squared personality you seek, without being too attached to the roman attitude.

The ‘s’ is much revised: I have rounded the curves and given it a bit more forward movement. I think the lack of squareness in the italic helps to differentiate it from the roman.


Hi Dan,

2 words: Much Better.
Glad to see you agree that you admit that the absence of squareness in the italic gives it its own personality. It becomes quite interesting to see how the two forms are so close and so separate at the same time.
Only one thing: it seems to me that the uppercase 'S' hasn´t got the same balance as you have given now to the lowercase one.
I liked the detail of the relationship between the uppercase 'T' and the lowercase 'h'. They seem to fit like a key to the lock. Still, as the spacing seems to be important for you, could you consider ligatures between uppercase/lowercase?

(No, it's not an obsession :D)


another thing, you won´t be mad at me if I tell you that I'm preferring the Italic best? It's not about achievement, but personal taste. Still, good work


Really nice work. What you've done with the ampersand italic and the french look you've gave to him it's perfect.

hello dan!

I've just noticed something a bit weird. I printed it in order to be sure that ismn´t a visulization problem. The 'e'. The stronger lower part makes him lower than the actual baseline made by other letters. I believe it's due to the strength of the the lower stroke, more than the height. Could you take a look at it? The italic version seems better, but the roman should be analized.

It's almost perfect. Keep up.


One more thing. The rendering of the lowercase 'a' in the italic version seems to have a problem with the rendering. Is it my problem?

View original article