In-Progress Final Project: Typeface Classifier Map

by Nara @ 10:04 am 19 April 2010

Progress Report:

PCA is working (using the library at

Doing lots of reading on shape detection

Will make interface a little more interactive as well

1 Comment

  1. Comments from Crit 4/21:


    This is really well-researched. I think you should write a paper. ATypI? There are conferences that would love this work. And designers/others in the Processing community would find this a valuable way of learning about PCA.

    History is such an important part of a typeface’s personality. Maybe you could provide some information about when the face was made?
    – Oh, you did it. Never mind.

    Time periods view is great. I think that the interactive component is really helpful in reaching an audience that the paper won’t, and being an even more impressive portfolio piece
    Will be really popular whenever online in a final form!
    Yes, I agree that time periods is very enlightening.

    It’s not entirely obvious how things are grouped. I know PCA is magical, but it would be nice to see the parameters you’re working from.
    I agree. What do the axis represent?
    Great project. Write a paper and I think it will help to explain all the work you’ve done to analyze the typefaces.

    Do all 15 different versions of Garamond! awesome suggestion from Patrick.
    yes: knowing what the actual components are would be interesting.

    Is there a better way to deal with the overlapping type faces? Maybe 3D?

    I want this to pick a font for me. Maybe it could pick out keywords from whatever you’re writing and choose the appropriate font for the context.

    Add lots o’ fonts.

    I first thought this project was blah, but I’ve changed my mind. I think it is a thought-provoking work that explores the world of typefaces in an interesting way.

    It might help to make the display a little easier to read…for exampe, make less common fonts a shade of grey and behind those that are more well known. Also, when you hover your mouse over a font, renderit on front in black at a slightly larger size. This way you can search through the fonts and select one more easily.

    I like the frog version more than the airplane; although, it may be nice to be able to choose between the interfaces. For the frog interface, maybe have different “levels”–let each level be determined by the speed of the flies. You may also want to try to have each interface be more related to each other. For instance, instead of a paper airpane, have a bird. Nice job, good luck! -Amanda

    If you were to write a paper, these are just some thoughts. There’s been a lot of study on typefaces, analyzing their features and so forth. But what you provide here is an analysis based on quantitative data (your metrics). So maybe the paper could focus on classifying the typefaces based on numerical data. When the typographers were creating their fonts, they probably didn’t worry too much about the metrics but rather the visual characteristics. Therefore, it’s interesting how the typefaces happen to sort themselves out when you use the metrics and I feel the analysis of these quantitative metrics versus their desired styles is interesting.

    Comment by golan — 25 April 2010 @ 1:46 pm

This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License.
(c) 2016 Special Topics in Interactive Art & Computational Design | powered by WordPress with Barecity