This is an experiment from 2011 in which the author produced a font by averaging all the fonts on their system.
I'm reposting it here because I noticed that this looks a lot like the uncanny valley produced when an image AI tries to make text, which makes perfect sense: it's a statistical average of fonts.
Same. It looks like the print you see in old books. Very pleasing to the eye. The lower case 'm' sticks out to me though, the second hump is raised a little too high.
Interesting how modern designers think readable fonts (with serifs, so people can reliably distinguish between Al and AI, for example) are "uncanny" because they don't follow the latest trends in ultra-minimalist "design" and other fashions.
Most serif fonts look too noisy to me, I made a website the other day and set a serif font and it immediately stood out to me as cluttered. I have zero design sense, though, so this is just an opinion.
I've used Averia (Serif Libre, specifically) for at least a decade as my primary font for email, web pages in 'reader' mode, writing long-form text, etc. I find it extremely legible, and even calming.
Ironically, I've been a typographer for decades, both for print and online. Averia might seem an odd choice for someone intimately familiar with typographic theory/history and the vast catalog of possible fonts. But there's a certain pleasure and comfort in a font that is not trying to stand out or do anything particularly special.
Does this font simply ... Not look good to anyone else? It is visually kind of uncomfortable to behold. Maybe it's because it's a bit blurry feeling.
It sort of suggests to me that there's a lot going on with typeface design that we take for granted.
Edit: on closer inspection, the letter forms are kind of all over the place. The humps on the 'm' are lopsided, letter heights are sort of random. I think it's an interesting idea but to make it a more useful font would take a lot of manual fine tuning.
It's because the website is using cufon, a very early attempt at supporting custom fonts on the web using HTML canvas - basically every word you see is rendered as an image rather than text. The end result does not look good on hi-dpi screens like modern Macbook displays, probably they did not exist back then. The site mentions Google Font has a hosted version of it now and you can look at how it is meant to be rendered https://fonts.google.com/specimen/Averia+Libre
I’m surprised by how good it looks. This is really cool! I do feel like the Q and 4 characters need a little manual tweaking since the blur+threshold technique leaves some artifacts in the corners but those are such minor issues given how readable this font is overall. Love it.
This is really cool. There's something very pleasing about precisely how unobtrusive it feels. You can also view the specifically serif-only and sans-serif-only versions here:
I think it would be really cool if a designer used these as a starting point for overall metrics, but then regularized and cleaned them up to exhibit consistent proportions and elements from character to character, without the wobbly parts. It really feels like it would become an ideal font family for reader mode, for journaling, just any time you want to focus on content and have a font that just "gets out of the way".
> I call it Avería – which is a Spanish word related to the root of the word ‘average’. It actually means mechanical breakdown or damage. This seemed curiously fitting, and I was assured by a Spanish friend-of-a-friend that “Avería is an incredibly beautiful word regardless of its meaning”. So that's nice.
Very cool project, thank you for sharing! To me, it raises some interesting questions around attribution of sources in derived works, in the same way that AI training does.
This is an experiment from 2011 in which the author produced a font by averaging all the fonts on their system.
I'm reposting it here because I noticed that this looks a lot like the uncanny valley produced when an image AI tries to make text, which makes perfect sense: it's a statistical average of fonts.
It also reminds me a bit of what text looks like after multiple rounds of photocopying. Like the handouts we'd get in grade school.
And the smell of weird, purple mimeograph[1] ink.
[1]: https://en.wikipedia.org/wiki/Mimeograph
I don’t get uncanny valley feel from this one. It feels kind of great for me as a font.
Same. It looks like the print you see in old books. Very pleasing to the eye. The lower case 'm' sticks out to me though, the second hump is raised a little too high.
Interestingly it evokes Open Dyslexic.
Yes, I saw the exact same thing when you posted it - "oh, AI text looks like an averaging of fonts".
I wonder if you can ask AI to use a particular font for text in generated images.
Interesting how modern designers think readable fonts (with serifs, so people can reliably distinguish between Al and AI, for example) are "uncanny" because they don't follow the latest trends in ultra-minimalist "design" and other fashions.
Most serif fonts look too noisy to me, I made a website the other day and set a serif font and it immediately stood out to me as cluttered. I have zero design sense, though, so this is just an opinion.
I like readable serif fonts but this one really looks like an uncanny AI image.
I've used Averia (Serif Libre, specifically) for at least a decade as my primary font for email, web pages in 'reader' mode, writing long-form text, etc. I find it extremely legible, and even calming.
Ironically, I've been a typographer for decades, both for print and online. Averia might seem an odd choice for someone intimately familiar with typographic theory/history and the vast catalog of possible fonts. But there's a certain pleasure and comfort in a font that is not trying to stand out or do anything particularly special.
It's kind of like how if you take the average of enough male or female human faces, the result is a very pleasing, attractive face.
Same with music, a large group of people singing slightly off-key (each in their own way) tends to sound pretty good in aggregate
Does this font simply ... Not look good to anyone else? It is visually kind of uncomfortable to behold. Maybe it's because it's a bit blurry feeling.
It sort of suggests to me that there's a lot going on with typeface design that we take for granted.
Edit: on closer inspection, the letter forms are kind of all over the place. The humps on the 'm' are lopsided, letter heights are sort of random. I think it's an interesting idea but to make it a more useful font would take a lot of manual fine tuning.
It's because the website is using cufon, a very early attempt at supporting custom fonts on the web using HTML canvas - basically every word you see is rendered as an image rather than text. The end result does not look good on hi-dpi screens like modern Macbook displays, probably they did not exist back then. The site mentions Google Font has a hosted version of it now and you can look at how it is meant to be rendered https://fonts.google.com/specimen/Averia+Libre
Wow, that looks completely different than how Safari rendered the site. Thanks for the link. I like the look of it hosted at Google.
Aaahhh, thank you.
It still looks a little funky but way more readable.
I’m surprised by how good it looks. This is really cool! I do feel like the Q and 4 characters need a little manual tweaking since the blur+threshold technique leaves some artifacts in the corners but those are such minor issues given how readable this font is overall. Love it.
This is really cool. There's something very pleasing about precisely how unobtrusive it feels. You can also view the specifically serif-only and sans-serif-only versions here:
http://iotic.com/averia/preview.php
I think it would be really cool if a designer used these as a starting point for overall metrics, but then regularized and cleaned them up to exhibit consistent proportions and elements from character to character, without the wobbly parts. It really feels like it would become an ideal font family for reader mode, for journaling, just any time you want to focus on content and have a font that just "gets out of the way".
Btw, "Avería" means "failure" in spanish
Arabic ʕawāriyya for (goods) damaged in transit > Catalan avaria for a breakdown, damage > Spanish avería for a breakdown, something that has failed
https://en.wiktionary.org/wiki/avería
This is mentioned:
> I call it Avería – which is a Spanish word related to the root of the word ‘average’. It actually means mechanical breakdown or damage. This seemed curiously fitting, and I was assured by a Spanish friend-of-a-friend that “Avería is an incredibly beautiful word regardless of its meaning”. So that's nice.
"Average" comes from Arabic for "damaged goods."
I kind of dig this. It seems like it might look good on an ereader. Might have to upload it to my kobo!
Very cool project, thank you for sharing! To me, it raises some interesting questions around attribution of sources in derived works, in the same way that AI training does.
We already have the average font and it’s the execrable Lato.
https://fonts.google.com/specimen/Lato
https://fonts.google.com/?query=Averia
I'd love to see the results for the same process used on monospace fonts.
Looks blurry on my phone.
The site uses bitmap images, not web fonts.