New Years Resolution - A typface diet

Moseph's picture

Hello Everyone,

Looking for some input on a New Years resolution I'm making. For the next year I want to work with a limited set of typefaces.

The idea of working with "a few good typefaces" has always appealed to me, but I've never decided what those should be. Now that web fonts are sufficiently supported (90% of my work is web design), I have decided to pick 20 typefaces to work with all year.

But I have a slight problem. I narrowed my selection down to 17 typefaces and don't really see any more that seem necessary. Do you see some glaring omissions in this list? Or is 20 not really a very restrictive number? Should I aim for 10, 5?

Here's the selections so far:

1 - Beufort
2 - Caecelia
3 - Clarendon
4 - Chapparrel
5 - Palantino Nova
6 - Perpetua
7 - Georgia
8 - Iowan Oldstyle
9 - ITC Charter

1 - Avenir Next
2 - FF OCR
3 - Gill Sans
4 - Myriad
5 - Optima Nova
6 - Sweet Sans
7 - Trade Gothic
8 - Verdana

What would you add or subtract?


hrant's picture

I'm sorry, but I think that's a stinky resolution. The good news is you have plenty of time to find a good one. Like: resolve to make a font. Take the whole year.


Moseph's picture

Hah, yeah, that would be a far more noble pursuit! But I doubt my type design skills would fetch much money!

Care to elaborate on why the resolution is "stinky" in your opinion?


hrant's picture

I didn't realize your resolution was about money. So using fewer fonts brings in money? Are you going to buy all those fonts in advance?

It's stinky because it's artificial - it misunderstands what a font does. Good font selection depends on the project, and the only way a canned list could make sense is if you can predict the future.

You know Massimo "Six Fonts" Vignelli? Worse than the worst font pirate. At least the pirates appreciate fonts. I remember with fondness the time I confronted him on the ATypI list.


HVB's picture

It depends on what you do.

If all you do is send emails, then you probably have 16 too many.

If you create posters and advertising for clients, then you have thousands too few.

- Herb

Moseph's picture

Thanks for the feedback. I can see your point(s). The sites and companies I work for are in conservative industries (utilities, insurance, law, architecture). So I guess I am able to predict the future to some extent.

As for saving/making money, these font's are available via my subscription for use on the web. If I use them in print, the customer is billed for them, so it's not really a profit thing, except I am hoping it will save me time in selecting typefaces (on average I spend 3-6 hours selecting typefaces every time one needs choosing).


Chris Dean's picture

From my position, I think the entire serif list can be scraped, and on the sans side, the only one worth keeping is Gill. I’ll give it 2 mins and the rest of this pint for now:

Univers, Gill, Futura.

Garamond, Baskerville, Bodoni, Caslon, Century Schoolbook, Courier, Starling.

And in all honesty, I could do 9o% of my work with just two. It’s not the font, but how you use it. A skilled typographer can set a wedding invitation in Helvetica. Or, as my Grandfather once told me, “Tis a poor craftsman who blames his tools.”

hrant's picture

'Tis a poor craftsman who uses a hammer to drive a screw.

Anybody who can't see that each job has its own ideal font doesn't grasp the first thing about type.


Chris Dean's picture

I’m not denying that every context can use a different face, I’m just saying that if push came to shove, through use of colour, size, weight, space, &c, I could solve most any problem with Times & Arial. And I know you could too.

hrant's picture

We must mean different things by "solve". And remind me not to let you design any of my kids' wedding invitations.


Chris Dean's picture

remind me not to let you design any of my kids' wedding invitations.”

SNAP! Always nice to end the day with a smile ;)

Moseph's picture

Thanks Dean, I appreciate your viewpoint. Thanks for the link to Starling as well, I'll have to take a good look at that.

All of the other fonts you mentioned were on my short list and got removed for the following reasons:

Univers - too similar to Myriad on screen
Futura - x-height is a bit small for screen. Replaced with Gill Sans

Garamond - replaced with Iowan Oldstyle
Baskerville - Replaced with Georgia (on the screen, pretty similar)
Bodoni - Removed. Too difficult to reproduce on screen at even medium sizes
Caslon - Removed. Too similar to Iowan Oldstyle on the screen
Courier - Completely overlooked selecting a mono-spaced font, thanks!
Century Schoolbook - Replaced this with Georgia, but that might have been hasty, on screen they're fairly distinct.

Please don't cringe at these substitutions, I do realize these typefaces in print are very distinct from one another and have unique histories. But since this exercise is almost exclusively going to be carried out on the screen, these typefaces become similar enough (to my eye at least) that I didn't think the redundancy was warranted.

Is this way off base? Are there better compromises to take, or important features that I missed?


Chris Dean's picture

I did not realize they were strictly for screen. Given that we are rapidly approaching resolutiution that exceeds our visyal acuity threshold, fonts designed for screen are irrelevant. Basically, past a certain point, we can only see so many dots per inch. Some say we are there with apple’s Retina Displays. As such, designing for screen is finally moot.

Regarding your choices, I still think you’re crazy ;).

First, Gill is sooooo far from Futura there’s no way one can possibly replace the other. One is humanistic, looks good as a sans for small body copy, the other is almost perfectly geometric. Just compare the [a] and [t] glyphs. The Berthold Futura also comes with small caps and oldstyle figures (SC & OSF). Rare for a sans.

As far as your other replacements, Myriad = Univers, they don’t even come close to resembling one another, and Univers kicks the siht out of Myriad bassed on the size of its family alone.

Bodoni is a tough call, and I might give you that one. I have never used it for body copy, and sometimes for display. But in an “I can only have so many” list, I think having a didot is essential.

Caslon. Possibly one of the best workhorse fonts even designed. “When in doubt, set it in Caslon.” Somewhat of a “if theres a chance things are going to go totally of the rails, I better use this one just in case.” However, this adage came from the days of potentally poor paper and printing (deliberate alliteration).

Century Schoolbook, again, not even close to Georgia so I’m not sure where the replacement logic comes into play. Century is another fantastic workhorse. Easy to read (some say bassed on frequency of exposure), good x-height, strong bold, salient italic, another very good go-to. Personally, I prefer ITC Century.

Oh, and I may have missed a few. You should have a glyphic, Optima is pretty standard, but has quite a small family. I prefer Nick Shin’s Beaufort. Mainly due to the size of the family, and the presene of SC & OSF.

Also a blackletter. Possibly San Marco as it traditionally pairs well with Garamond. I think that came from Bringhurst, but my library is in storage. If you went for some Goudys, you’d be killing two birds with one stone (serif & blackletter from the same family).

For a swash I’d have to go with Poetica again, due to the size of the family. That, and the set of ampersands simply kicks ass.

Script, never used that many, but I find myself leaning towards Kuenstler.

When selecting a typeface, two of the most important things I consider are:

1. The size of the family (range of weights, italics &c).

2. Does it have SC & OSF?

If it doesn’t meet those criteria, it seldom makes it to the list of contenders. If you *cough*Paul*cough* want to get into a discussion about appropriateness for the context, I will still say a skilled typographer can use most any typeface and have it speak in a certain tone of voice through use of additional design decisions such as colour, space, size &c. The same old Helvetica wedding defence.

And definitely look into Starling. It truly is a work of art (and Mike Parker is a wonderful person).

I still might be missing your intention, but it seems as though most of your selections are cheep, if not free, and lousy covers of the real thing. Right down to being pre-installed system fonts. Is there a logic to this?


Moseph's picture

Thanks Chris for your input. A couple of contextual clarifications.

1) Based on my clients web analytics, less than 5% of their customers are using Macs. So very few of them will have retina displays (and honestly, I think we're still several years away from these displays being mainstream)

2) The typefaces that were selected which also happen to be system fonts were chosen because they have been hinted for the screen. Honestly, I don't know if many of the other typefaces recommended here are. Until retina displays are mainstream, this is still important for anyone serious about type looking crispy on the web. Sadly, a lot of great typefaces designed for print turn to absolute mud on the screen.

But as for using Helvetica on wedding invitations, I couldn't agree more! Used it last year for a baby shower card. Worked great.


hrant's picture

Plus "retina" displays being sharp enough to make pixels irrelevant is bogus. For one thing I can still see the gray miasma on Apple's rendering anywhere, and I'm 44.

BTW hopefully that baby won't turn out to be a competent designer, see the card, and hate his parents.


Chris Dean's picture

You really should post your Helvetica/wedding/baby comments in the Typography jokes thread. Thats two really good zingers so far.

Moseph's picture

Haha, the odds aren't too bad for him becoming a designer. His mom's an architect and his dad's a web developer. But it might have the opposite effect; I remember asking my friend why he was so fond of Futura and he said 'My favorite series of story books as a kid was set in futura, I looked at those letter forms for years, so now they always make me feel good." So perhaps the new baby will just associate Helvetica with the love he received as a child (like one of my clients who calls it the "Sesame Street font!")


Ryan Maelhorn's picture

Personally I think a wedding invite done in Helvetica could be awesome.

Joshua Langman's picture

In fact, Michael Place talks about setting his wedding invitations in Helvetica in the Helvetica film, and mentions his wife vetoing a colophon.

Joshua Langman's picture

And, to actually address the topic of the thread, I disagree quite a bit with your list — of course every designer will disagree with a list like this, but here's my opinions:

Chaparral (note the spelling) is fairly similar to Caecilia — both kind of humanistic slab serifs. I would never use Perpetua for anything on screen. It's too delicate. What about some classic serifs — a Garamond, a Jensen, a Janson?

Baskerville is very different from Georgia — I'm not sure how that replacement works. Georgia is more similar to Caslon than Baskerville.

There's certainly a more interesting monospaced font than Courier.

How about Scala, which looks great on screen?

I could do without Verdana, and I might add Scala Sans or some other humanist sans.

If you think through the different categories of serifs and sans fonts, and different historical models, you can probably put together a more comprehensive list without making it any longer.

And a few nice display faces are in order too....

Here's a thought — since this is a resolution, why not resolve to license and use only recent releases for your list of fonts? So for every category, choose one new font by a smallish or up-and-coming foundry to use. Then you would be limiting yourself to your list, but also expanding your knowledge of contemporary type design. And you could get to know some lesser-known fonts and support some independent designers along the way!

Good luck!

Moseph's picture

Wow Joshua, thanks for your response! A lot of good ideas and information there.

The responses to this thread has really made me rethink this whole idea. A lot of the typefaces I chose are not really appropriate for the times of type on the web. That said, you're idea of only using new (and I would add, screen optimized) typefaces from up-and-coming foundries would be a better resolution.

Thanks everyone for your blunt-yet-reasonable insight.


charles ellertson's picture

Asking type designers about using fewer fonts is sort of fruitless. They tend see it as a loss of income.

Most of the really good book designers I know work with about 4 text faces. They'll change those from time to time. A few more for display. The plus is they get to know the fonts they use very well. One exception was Richard Eckersley, he'd usually cut that to one or two fonts. Believe he only used Trinite (and Baskerville, if Trinite was just wrong) the last 3-4 years of his life.

The paper used for printing is a factor -- if this is indeed about print projects. Hard to find fonts that work equally well on both coated & uncoated stocks.

And BTW, for text, Perpetua doesn't work with ink on paper, either, unless you're talking 14-point, metal type/letterpress printing. For display, it can work. However, I suspect from the fonts you've selected you're doing work for the web?

hrant's picture

Even if I won the lottery (unlikely since I don't play) I would believe that limiting your font choice is merely compensation for some other shortcoming.


Moseph's picture

hrant, what I lack is what Charles_E referred to, the deep knowledge of a few really good typefaces. Right now I have a broad knowledge of a broad set of typefaces. What I was hoping to achieve was a way to force myself to gain a deeper understanding of the type I work with. However I'm quickly learning that this will not be as easy as I'd thought. Because while great typefaces for print are well known and well critiqued, type for web (apart from system fonts) is in its infancy and licensing is all over the place. As it stands now I'll be lucky to work with fonts even half as good as what have been developed for print, at least for the next few years.

I'm appreciative for everyone's input, I realize now my potential new-years resolution was off base. I'm formulating a new one now which will hopefully get me where I want to be.

Thanks again for all the great feedback!


hrant's picture

Admitting it's hard to really know a typeface is great. The problem is claiming that limiting them is a Good Thing, instead of realizing it might be a necessary evil. An even bigger problem is saying that everybody needs to use few typefaces.


J. Tillman's picture

Charles_e, I know I'm going off topic, but could you offer a brief explanation of why some fonts work only for coated or for uncoated paper? Thank you.

Edit: The fonts that look good on the laser printer beside my desk, which category would they be in?

Nick Shinn's picture

Asking type designers about using fewer fonts is sort of fruitless. They tend see it as a loss of income.

Before I became a type designer I was an art director of the “so many flavors, so little time” persuasion.
So my taste for variety has nothing to do with income.
(Blame the Beatles for indoctrinating me with eclecticism.)

charles ellertson's picture

Well, any reply is going to get disagreement, but I'm a big boy, so here's an example.

Far as I'm concerned, Merlo work fine on a coated stock. But not on uncoated, if set large than 10 point, and at 10-point, it is hard to read for anyone of the eyeglass generation.

Same holds true with a number of Adobe's medium weights. Far as I'm concerned, the "regular" weights in .otf were set up for best appearance when laser printers were used to output the type. Probably not a bad decision on Adobe's part. For book fonts, Adobe assumed coated stock, and gave us the medium weights. But they're usually too heavy when used on uncoated sheets, where the ink spreads more.

It really is a matter of how the ink is applied to and taken by the paper.

* * *

Type that looks good on a laser printer will be a bit weak if printed offset, on uncoated paper, and 2400+ dpi with direct-to-plate plates. If you go back to when PostScript was first printed without an intervening repro stage, and again when direct-to-plate technology came out, you'll see all sorts of complaints about type being too thin & spindly, even though it looked good on lasers. Usually the printer was blamed, but it wasn't the printer's fault, it was the type publishers. Remember that with DTP, the halftones are in the same file, and dotgain has to be minimized. Simply increasing the exposure when burning the plate isn't an option for the printer.

That's an old argument, and the counterargument today is everybody's eyes have adjusted to thinner, weaker type. Not a very good counterargument, IMSLTHO.

* * *

As for "understanding" a typeface, Rich Hendel remarked that with Trinite, he could steal Richard Eckersley text specifications, item for item, and the book still wouldn't look right. I know what he means, and can't explain it, either.

Broadening that to a more reasonable position, there will be certain measure/size/leading/margin relationships that work better than others with individual types. Varying one does NOT mean a simple scaling of the others. When you go from a 6-1/8 x 9-1/4 to a 6x9, or have to get a few more/less characters on a page, what do you do? That's where having a type in your fingers pays off.


I do know a couple book designers who could just take a font & make it work. During the 1980s and 1990s, the designer who got the most books in the AAUP Book Show for any given year was usually either Richard Eckersley, Rich Hendel, or Mary Mendell (an alphabetical sorting). Mary could, and did, make just about anything work well.

dickface's picture

The only typeface you will ever need

Chris Dean's picture


Syndicate content Syndicate content