Thanks for the feedback, and thanks to hrant for startle people up. I have no pretension to feedback, as I can give back so little at the moment...
Here is PDF V25.
I've considered your comments, and those I got for Sofa Sans (the ones that are applicable on serif as well). Italic, Sans and Antiqua is is put on ice at the moment, as I want to finish this first.
bdpq: thinned bowels a bit. they were too heavy in print.
g: slightly less weight, based on my printing results.
k: lower part slightly more heavy (printing results).
r: narrower, beak. what do you think about it? first real life feedback was positive.
s: a bit wider and lighter. But I think it's too gaunt now. Difficult...
t: less hooked, you are absolutly right, this was inconsistent. Maybe I have to open it even more.
u: slightly wider.
vwxy: shorter serifs. goodbye elegance, welcome practice...
Q: You are right, Q was a problem I have ignored yet. there are 4 alternates now. see the image below for details. I like Number 3 in the image best.
ae: smoother now
eth: bigger (based on my printouts, it turned out to be too small)
ring: more weight
breve: more weight
2: a bit wider
4: goodbye serifs... Should I kill the bottom one as well?
5: stroke weight redesigned
Hrant: what were your concerns about OS-numerals?
Randy: Thanks for the Robofab-Link. I try to get this working, but I'm no programming crack, so this could take some time :)
> what were your concerns about OS-numerals?
Their up/down distribution, which is unique. Unique isn't
necessarily bad (depending not least on how much text the
design is intended for) but Unique does benefit from some
sort of justification (in a craft).
Hrant, I think you may have misinterpreted the sample on page 3, which shows five os numerals and five lining numerals. Both sets are on the last page. I must admit, though, that the way you saw it does have some intriguing potential. A numeric progression from small to descenders to ascenders would mean that you would get a vague idea of how big a number is with only a glance. That alone wouldn't necessarily be worth flaunting convention, though.
Oh. Sorry. :-/
> you would get a vague idea of how big a number is
Hadn't thought of that! Sort of how the visual density of
a glyph in Hangul indicates its grammatical function,
providing an extra layer of information for the reader.
> That alone wouldn’t necessarily be worth flaunting convention, though.
I dunno. The conventional scheme is not only partly arbitrary, but it has problems too, most of all that the overall height position is too low compared to the lc. The old French OS scheme -with 3 and 5 ascending- is better in my view, and readers don't consciously notice the difference, so it probably has a "subvisible" functional advantage. In addition, one could argue for an acending 2 as well.
See also:http://typophile.com/node/15443 _http://typophile.com/node/2253 _http://typophile.com/node/7701 (Looks like a monologue because anonymous posts have been supressed.)
Numeral design really has superb potential for growth.
Argh, sorry... my fault.
I will correct this in the next PDF :)
The third Q is definitely better (and suited to the whole).
I would drop the others. Other solutions (like the 2nd), although initially original, have been a little abused recently.
Besides, as a general remark, I think the face would benefit from a general spacing tighten-up. At least the Roman.
Lots of finetuning, optical tweaking, kerning, some opentype-scripting, etc.
How do you like the new "4"?
compare it directly to V_25 on Page 9 (black=new, red=old)
New: Smallcaps-Versions of currency symbols and some others. Some of them still have wrong stroke weight though.
Are there any other glyphs left that could need a smallcaps-version too?
Thanks to this thread I'm motivated again. Want to come to an end with it soon, then polish the italic, do a lot of beta-testing, publish it and then getting RICH RICH RICH (or die tryin' ;-).
I love this so far. It truly is a "cozy" face. It reminds of something Underware would do, actually. I'd certainly purchase this.
Sorry to take so long to make comments. I have to budget the time I give for online critiques; your fine craftsmanship and thoroughness tell me you intend to develop this fully. Very nice specimen texts as well.
I still see some big kinks in the basic set; with a design like this whose modelling is so subtle, these stand out and snag the eye:
* Center crossbars on EF too short. Serifs in these letters do tend to clash or close up, but see Dante for an example of a successful application of long serifs in these shapes.
* Joint of diagonals of K with vertical very dark and crowded, makes letter look narrow. Rearrange where these strokes meet.
* Some serifs are still much too long; see AVWY and tails of K and R. Conversely, the serifs on C, E, L, T and Z are unnecessarily short.
* The obvious humanist proportions of the lowercase call for corresponding variation in width proportions in the uppercase. Homogenizing the UC widths creates a deadening feel which doesn't fit the lowercase. See E, F, T (wide) and K, M (narrow). A will look better if the outside serifs are shortened.
* Do the R and K need to have a drooping, bowed leg? Elsewhere in the design, stems seem rigid (in a pleasing way) and these sagging elements seem like they were left out in the sun. Consider reversing the bow, for more pleasing tension.
* The nice tapered serifs and the tapered stems imply a range of contrast that the general stroke weights don't deliver except in a few letters: d, e, h, r, z. Consider bringing the contrast modeling, with the color benefits it brings, to the rest of the letters. Especially caps like B, D, E, F, G, M, N, O, R, S, W will benefit from some lightening in appropriate spots. Also a, g, k, w and z will improve with some adjustment in contrast. A striking example is how monoline W and S look compared to d and n. Look at S and T together.
* Why does the y have such an emphatic tail? The bend at the end of the stroke is already eye-catching; adding a serif at an abrupt angle makes it inescapable. The angle and density of the serifs on s and S also act as snags to the reader's eye. Consider at least changing the angle of these.
* Apply the corner-lightening trick more in letters like m, n, u, r. the odd horizontal serifs on those letters add weight at the x-height that is continued in their monoline arches and other serifs like those of vwy. LC a would really improve with some thinning of joints.
* The i-t ligature should be removed: When I read your sample text, it appears exactly to my eye as an f-i ligature. Consider removing other ligatures which distort the components beyond readable recognition: cr, cl, sl. Also consider eliminating ligatures that aren't necessary; is reading enhanced by an sk ligature? It isn't strictly an improvement.
* The oldstyle figures look preliminary. Contrast, proportions, shapes, and especially the 4 look thrown together. I realize you've put a lot of attention on the 4, but it doesn't even appear to be from the same font. Not because of the unconventional shape, but because it's still so monoline, and appears to be leaning and bent. Balance is what it needs. 5 also looks like a transplant from a grotesque sans. Look at other typefaces to get ideas about where to apply weight in the shapes of 2, 5, and 7.
* The weight issue of g can be solved with *redistribution* of weight, rather than simply making it lighter overall. This shape, because of its complexity, needs this kind of modeling anyway, especially in a Slab or Sans design where horizontal stems pile up and make a letter dark. You see this in a, e, g, s, and B.
The spacing in the text samples is very loose, especially considering the narrow space character. Is that text tracked? I think the letters need closer fitting overall. Put two m's together to see what I mean.
Many of my comments seem to ignore the "program" or "logic" of this design. If there is any rule I think worth following in type design, it is: Rules of logic fall apart more and more as a typeface is realized, and the result should only appear to the untrained observer as if serifs are actually the same, stems are identical, etc. In this design, I see that the sharp serif angle and low contrast are apparent themes. However, where they impede smooth reading or distract the eye, they are only mannerisms.
Other than these comments, I see the potential for this design to be very useful in many settings. It does convey the comfortable quality you intend. Eliminate the few scratchy, lumpy elements, and it will be a chaise worthy of weekly afternoon naps for years to come....
Ooooh, Carl ups the ante. Very nice.
Wow, thanks a lot for taking the time for this feedback, Carl! These are the details that I missed so far, without knowing where to get them from.
I will need a dictionary to understand everything and and it will take me a lot of time to introduce the improvements beside my job, but this is very appreciated and it will not be forgotten.
I'd buy it the minute it was released.
So version 40 is up (see first post).
There are lots of changes suggested by crossgrove, though not all yet.
Concerning ligatures: The using in small sizes in the sample sheet actually is "wrong" (respectively only for testing the spacing of them). I am aware that they should only be used (with deliberation) for display purposes.
Concerning loose spacing: The sample is not tracked, the setting is that loose at the moment, which might be a relict of even longer serifes. I'm considering to tighten it a bit, though I'm not sure to which extent: in Europe, typographers tend to prefer a bit looser spacing than in America, if my perception is right. I am influenced by that obviously, but anyway, spacing is too loose at the moment.
Concerning program of a font: I am aware of, that one has to leave the path of strict guidelines when giving a design perfection. But as a beginner, this can be dangerous: I'm afraid of getting stuck in small optical changes and lose track of things. This is why I tried to avoid as much of "irrational" tweaks as possible until now, but I know that this keeps me from finishing it. I have to grope my way piece by piece now :)
Thanks for your encouraging and detailed comments. Without this feedback I'd be stuck in this state of work...
Oscar: thanks for your comment, this tells me that I'm not doing this all just to file the whole thing in the end...
I specially like the ornamental/floral works on your homepage (that doesn't mean that I don't like the rest...)
Greetings from Austria
Sebastian, I'm glad my comments were useful to you. Sometimes a design gets stuck in a dead end and it's hard to back it out without a fresh look.
"should only be used (with deliberation) for display purposes"
I hope you are ready for all the horrible things people will do with your font that they shouldn't.... Stretching, condensing, skewing, outlining, tracking, hitting the "punk" filter in Illustrator.... and of course turning on all ligatures and setting a book. ;)
I see the improvements you've made. It's stronger already. Caps look much more settled. Your work is very subtle and decisive. The spacing, I think, is still too loose for general use, even if the trend is toward loose spacing. This design is somewhat unique in that the serifs are not all equal to stem weights, and there's a lot of proportionality to the design, so spacing will be very tricky. When that's done thoughtfully, I'm sure you'll like it better. If your monitor is big enough, I suggest viewing spacing at 2 very different sizes, such as 16 point and 80 point. Some spacing decisions look right at one size that simply are bad at another. I'm sure you noticed that already.
I like it!
> Sometimes a design gets stuck in a dead end and it’s hard to back it out without a fresh look.
A few months ago I couldn't believe how some typedesigners could work on several fonts the same time. Now I know why :). That's why I'm working on a new one (working title "dark chocolate", sketch-stadium). A very early version can be seen in typebattle#12, and a newer one, if chosen, in typecon exhibition. It's a pity that it is the least-thought-about one I have to offer...
And I'm even thinking about getting Terra Nova back on track. When looking at that one now, I see about 200.000 improvements to be made :)
> I hope you are ready for all the horrible things people will do with your font that they shouldn’t…
At the moment, I'm arguing with a friend of mine (a graphics designer), because he wanted a bold version of Sofa/Canapé, and used an outline to achieve that, even without asking me to redraw the four glyphs he actually needed... It will be shown about 4*3m in size – Beautiful! :(
I'm so shy about that because I'm afraid that it will destroy the kerning time already invested. But that's my fault, if doing such things before having the letter shapes settled down... Just need some overcoming to get it started.
What I've learned is that typedesign is suffering, especially as a beginner :). And what is needed, but understandably rare, is feedback from people who really "know". Amsterdam, Reading and other schools are far from here, and the decision to go there is not easy, so I'm dependant on friendly people who like to help.
Don't do any kerning until the shapes and spacing are final! In fact, remove all your kerning (argghhh!) so you can look at just default spacing.
This isn't so bad; when all the shapes and default spacing are done, you'll discover much less kerning is necessary. Kerning is just to fix pairs that can't look good by default spacing, like LT.
"Typedesign is suffering" ; D
So, 13 weeks later, here is the update :)
I've been working on Sofa Serif Roman to get it almost finished. To achieve this, I made some detours to a Black and Ultralight version and to the Roman Italic, to explore more about the general shape principles (and of course I proped heavily on Carls comments). This was very useful for me, it made me have a new look at the principles of the design, and I think I could improve the Roman based on that experience. (You can see Black and Ultralight them compared to the roman in the end of the PDF. They are not perfect yet and won't be until Roman is finished.)
Then I returned to the Roman, and I won't stop for another typeface until it's finished. I did a lot of finetuning of proportions, stroke weight and details, and I think now I can say the basic character set is almost finished.Please be critical and tell me about every minor error you still find! I think there are no major ones anymore, but I'm willing to negotiate ;-)
I've completely re-done the metrics. Spacing is finished, kerning is in finetuning and correction phase, and yes Carl, of course you were right: there is less to do now than in my first attempt.
What is still to be done:
- redesign ligatures, swashes, etc.
- check whether smallcaps are too thin (I have the misgiving they are...)
- Kerning of Uppercase<-->Smallcaps
- finetuning of some figures
- have a look at the opentype classes and coding whether it's still appropriate and complete
- rearrange diacritics
So if you have the time and willing, please concentrate your comments on basic character set (A-Z, Asmall-Zsmall, a-z, 0-9 and basic ASCII characters).
Of course, everything else is apprecialted as well, but know that I'm still working on them, so comments too much into detail may be overtaken by my own corrections anyway... But don't feel hesitated to comment conceptional things or major mistakes. For example, I don't know much about diacritics, It's mainly looking at solutions others have done, but's I can't say "now everything is correct".
Thanks for your help
Short intermezzo: which lig-style do you prefer?
Middle option, I hope you weren't considering just crashing two letters together as an option.
The middle one. Although it might be worth trying an ironed out version of leftmost "crashed" one (with the beak of the "f" in there).
No, it was the example how no-lig would look like.
I've chosen the middle one, but have weakened the wavy effect slightly.
:) I'll try this, hrant (and be it just for fun).
So here is the next update (first posting, version 61).
I think my last "big" post from 14th October was sabotaged by the shortly after "intermezzo". What I said there is still up-to-date, but I worked on, especially on:
- redesign ligatures, swashes, etc.A
- thickened smallcaps slightly
- a lot more of kerning with quotation marks etc.
- some opentype stuff
- rearranged diacritics
I will go on, but it's mainly bugfixing now, so if you find some of them, you could help me a lot.
There is still an overlap in the cursive (alternate italic) lowercase "r" (branch over stem).
How is this coming along?
I really like it, especially the "organic" effect created by your stroke variations, the sofa builder is fun, I'm not too sure of your ultralight version but I love the black one, a good looking font that made me think of Bussigny Serif. Nice one!
Roman can be considered as finished. It is just waiting for the italic and maybe a bold to be released. I have done kyrillic and greek letters, but they won't be included in a first release. They are not good yet.
Italic is on it's way to be finished.
Black and Ultralight haven't moved yet.
Sofa Sans: A modified and improved version is being used in a children's book I've made a view weeks ago. It does not have the same glyph count as Serif at the moment.
I've also done some experiments with an
- Sofa Antiqua: quite okay, but far away from finished, Bussigny reminds me of it...
- Sofa Script (not quite okay, what I want to do needs a lot of opentype and glyph variants).
At the moment, I have 15 hours of work a day, so I don't have the time to go on. I will continue in about one and a half months. I have an urge to do that, but it is not possible right now...