lundi 10 novembre 2008


I started with a little audit regarding monospace fonts that are used for coding/programming.
I picked up 7 from the most used monospace fonts by UNIX programmers.

Most of the fonts used in coding have relatively similar proportions. It is generally of 4:5, though best results are given by a 5:5 square proportion. The problem i found with a 4:5 proportion is that the glyphs curves become completely obsolete when rasterized. This means that there is no recognition possible between the outline version and the rasterized version. It's like the font was loosing it's identity when displayed under 10pts. Here's a screenshot of the problematic (font displayed at 9 and 14pts).

I started out doing some basic drawing in a 5x9 pixel grid for the basic lettershapes.

Obviously, with the small number of possibilities offered by a 5x9 grid, the letter's ergonomy has a strong minimal, low-tech boring connotation… What i want to do now is test various possible interpretations for the curves. The criterium is that i want to keep a same identity in the outline font and in the bitmaped, rasterized basic version. I did some drawing with Pixen and hum… Gimp (right?) and made tests with an alphabet, to question the sequences of letters as well.This is where i got with the pixel grid. For the next steps, i'd like to be able to get into scripting. I can take advantage of the fact that the basic drawing is made of squares (modules) and try to make tests by changing the modules. As i am completely new to scripting, i just simulated an example of what i'd like to be able to do with scripting; for instance add simple curves where a gap is found between pixel grid.

That's it for today. More to come these days.
(if anyone could help for the scripting—software, i'd appreciate…)

1 commentaire:

Dave Crossland a dit…

You should check out Gustavo Ferreira's work with pixel fonts, he's done some great stuff!