Commentary

The Average Font

Stephen Coles on September 29, 2011

A writer from The Atlantic Wire contacted me to get my opinion on the Average Font that’s making the internet rounds. I don’t think there’s anything worth writing about.

It’s the sort of project that most designers have seen many times before they are even out of school. It would be more interesting if there was a theory or direction behind Moritz Resl’s approach, but his short description shows there wasn’t much thought put into it.

“This project shows what a font would look like if it consisted of all typefaces installed on my system. Every character from a to z is drawn using every single font with a low opacity. In total there are over 900 typefaces in my library. I didn’t exclude the ugly ones.”

Exploring the commonalities and differences between typefaces is intriguing (though others have overlaid fonts to make lovely images that work as art better than Resl’s), but without any controls in the experiment, nor any data about the method or what’s included, the only thing we learn from the result is that there is variety in type, enough to make something mostly but not entirely illegible. And we get a pretty video.

Something like Kai Bernau’s Neutral, a well-researched comparison of typefaces in search of the most “neutral” aspects, has much more value. I guess the images wouldn’t the draw the traffic that Average Font draws.

Update: Yet another exploration in typeface merging popped up last week. This time the result, Avería, is a working family of fonts. My reaction is essentially the same as I wrote above. It’s an arguably interesting experiment, but not a very useful one. The designer, Dan Sayers, sums it up himself in the first sentence of his description: “I am not a type designer.” If you need a serif in this vein, there are far more useful typefaces drawn by trained professionals from scratch — Tribute, Fabiol, FF Avance, Garaline, Galena, to name a few. Or, to put it in a more festive way:

“I toast that creation with a glass of my famous 725-wine punch.” — Jonathan Hoefler.

Above: Average Font by Moritz Resl and Face Variations by W. Bradford Paley

This post is sponsored by brochure printing and postcard printing.

9 Comments

Luke Dorny says:

09.30.11 at 8:22 am

This is awesome!
Stephen, I’m not sure that you intended it as such, but your brief review of the idea of new glyph overlays seems a bit disheartening.
I’m sure there can always be more complete research on topics we engage in, especially online where brevity seems to trump completeness.
I don’t yet consider myself an old man, but I’ve had the wonderful chance to become acquainted with some fine folk that despite their talent and experience have become grumpy old “they don’t do things the right way anymore” folk.
However, I appreciate your covering of the topic despite this, and hope that you might continue to share your thoughts on topics like this, as I find them intriguing.
I love this site.
Stephen Coles says:

09.30.11 at 9:52 am

Thanks for the nice words, Luke. I don’t mean to be disheartening. I only mean to suggest that we can do better.
The quest for the average font says:

10.31.11 at 1:11 pm

[...] Stephen Coles of Typographica points out, the quest to find an average font is not unexplored territory. In 2006 visualisation expert W. [...]
Dominik says:

11.13.11 at 5:19 am

If what you looking for is the average font, you will get an average result. Self-fulfilling prophecy, isn’t it?!
Design work starts when the exploration ends.
vernon adams says:

11.27.11 at 11:24 pm

You are not talking about an ‘average font’ here, you are talking about an ‘averaged font’. There’s a big difference :)
Nick Shinn says:

12.04.11 at 2:17 pm

As an alternative to blah averaging by interpolation, perhaps it is possible to mix fonts by interbreeding them “genetically”, considering qualities such as serifs and contrast as traits that are dominant or regressive.
Eduardo Rodríguez Tunni says:

12.19.11 at 6:00 am

I did a similar project in 2005. The result was a typeface for text called “Average” which was later named “Media”. If you are interested I can show the work.
a r b o says:

04.10.12 at 4:45 pm

I can only agree with the leading view expressed here. It’s exceptionally hard to see the point. No wait, I’m being too kind. I don’t think there is any significance to the project.

Nick, FontLab can already blend 2 fonts together, so interbreeding by machine is here. But surely interbreeding by a human operator equiped with typographic knowledge and a vision will produce the most interesting offspring.
Miguel Catopodis says:

11.20.12 at 8:52 am

Regarding this topic, this “Average” project is very interesting.