4 ms·

Why didn't the designer select dot positions based on letter frequency or letter pairing frequency in English? This could have yielded a font that would be more

Why didn't the designer select dot positions based on letter frequency or letter pairing frequency in English? This could have yielded a font that would be more likely to make pretty geometric shapes. Even a simple improvement like vowels all having higher dot positions in the font, and consonants having lower positions, could be interesting. It seems that the dot positions are just based on letter position in the alphabet. There is probably a way to come up with dot positions that yield much nicer looking shapes.

jpfed 8y ago
My suspicion is that the way to improve legibility is to consider ways to make symbol images more robust to noise. In particular, if a symbol image were perturbed, would the context help you figure out what that symbol should be? For each letter, we can build a histogram of contexts that letter occurs in for a sufficiently large corpus. An example is the letter "o", whose occurrences in the previous sentence have the following contexts: f_r (twice), t_g, c_n, _c, c_r. Each letter's context histograms have some similarity with each other letter's. We want different letters that have similar contexts to have widely differing appearances, because if their appearances were similar, their contexts will be of less help to the reader in disambiguating. And letters with sufficiently different contexts may be allowed to be more similar in appearance (note e.g. the archaic long "s", which despite visual ſimilarity to "f" is easily distinguished from it by the different contexts it appears in).
- blueblimp 8y ago
  That's a very interesting point and suggests that, counterintuitively, it would actually be bad for (for example) vowels to look similar to each other. Another way to use language statistics would be to make the font variable-width (which may help with legibility) but make commonly-occurring letters thin.
- benj111 8y ago
  This doesn't preclude optimising vowels for example though, just them looking similar. I'm not convinced I could differentiate M N and O without context, having a known vowel to differentiate the heights you're looking at would help?
Kiro 8y ago
Because there are other languages than English.
- benj111 8y ago
  Dvorak (for example) is an attempt to optimise for english text, so it is worth it at some level. I have no idea how applicable it is to other languages.
- Koshkin 8y ago
  Ideally, each language should have its own alphabet/writing system that is optimized for the given language.
trogdoro 8y ago
> There is probably a way to come up with dot positions that yield much nicer looking shapes. Try it yourself! http://dotsies.org/design-your-own http://dotsies.org/design-your-own Hit the "random" link a few times and you'll get an idea of the possible variations. In the Dotsies pattern, vowels are at the tops and bottoms of words. The variations can look grainy, chunky, loopy, checker-board-ish (you'll see what I mean if you try it). The version I picked is a balance that leans toward being sparse and loopy, and is one of the many variations that has an intuitive letter ordering. Here's a decent one, but with a crappy ordering. http://dotsies.org/design-your-own/#comeyhangrxdtwqpsfubzljvik http://dotsies.org/design-your-own/#comeyhangrxdtwqpsfubzljv... Surprisingly, none of them jump out as significantly better than the others (the overly dense ones being exceptions). [from another comment] > I am pretty sceptical that an "optimized" alphabet glyph would be one dot for a b c d e, two adjacent dots for f g h i, etc. There are other mappings that looked marginally better to me, but I deemed picking one with a nice ordering was worth doing. A and E happen to be the 1st and 3rd most common letters, which works out if you want to lean toward the sparse end. Also O (the 4th), A, and E all stick to the top and bottom edges of words, so if you want your words to tend to have holes in the middle they are a good fit. Getting a decent mapping for the commonest letters seems to be key, but by all means, try the above link and post an alternative that you think is better.

jpfed 8y ago

My suspicion is that the way to improve legibility is to consider ways to make symbol images more robust to noise. In particular, if a symbol image were perturbed, would the context help you figure out what that symbol should be? For each letter, we can build a histogram of contexts that letter occurs in for a sufficiently large corpus. An example is the letter "o", whose occurrences in the previous sentence have the following contexts: f_r (twice), t_g, c_n, _c, c_r. Each letter's context histograms have some similarity with each other letter's. We want different letters that have similar contexts to have widely differing appearances, because if their appearances were similar, their contexts will be of less help to the reader in disambiguating. And letters with sufficiently different contexts may be allowed to be more similar in appearance (note e.g. the archaic long "s", which despite visual ſimilarity to "f" is easily distinguished from it by the different contexts it appears in).

blueblimp 8y ago
That's a very interesting point and suggests that, counterintuitively, it would actually be bad for (for example) vowels to look similar to each other. Another way to use language statistics would be to make the font variable-width (which may help with legibility) but make commonly-occurring letters thin.
benj111 8y ago
This doesn't preclude optimising vowels for example though, just them looking similar. I'm not convinced I could differentiate M N and O without context, having a known vowel to differentiate the heights you're looking at would help?

Kiro 8y ago

Because there are other languages than English.

benj111 8y ago
Dvorak (for example) is an attempt to optimise for english text, so it is worth it at some level. I have no idea how applicable it is to other languages.
Koshkin 8y ago
Ideally, each language should have its own alphabet/writing system that is optimized for the given language.

trogdoro 8y ago

> There is probably a way to come up with dot positions that yield much nicer looking shapes. Try it yourself! http://dotsies.org/design-your-own http://dotsies.org/design-your-own Hit the "random" link a few times and you'll get an idea of the possible variations. In the Dotsies pattern, vowels are at the tops and bottoms of words. The variations can look grainy, chunky, loopy, checker-board-ish (you'll see what I mean if you try it). The version I picked is a balance that leans toward being sparse and loopy, and is one of the many variations that has an intuitive letter ordering. Here's a decent one, but with a crappy ordering. http://dotsies.org/design-your-own/#comeyhangrxdtwqpsfubzljvik http://dotsies.org/design-your-own/#comeyhangrxdtwqpsfubzljv... Surprisingly, none of them jump out as significantly better than the others (the overly dense ones being exceptions). [from another comment] > I am pretty sceptical that an "optimized" alphabet glyph would be one dot for a b c d e, two adjacent dots for f g h i, etc. There are other mappings that looked marginally better to me, but I deemed picking one with a nice ordering was worth doing. A and E happen to be the 1st and 3rd most common letters, which works out if you want to lean toward the sparse end. Also O (the 4th), A, and E all stick to the top and bottom edges of words, so if you want your words to tend to have holes in the middle they are a good fit. Getting a decent mapping for the commonest letters seems to be key, but by all means, try the above link and post an alternative that you think is better.