How's my pixely Greek?

This is the base of a pixelated font, for which I'm considering adding Greek. Happy to hear any feedback.


  • Could easily use work: δ σ ω. Would rather change the μ design a little too but it's reasonable. Greek reader, not native.

    Query also how you would even do polytonic at this scale (A: you wouldn't). You definitely need to have a tonos though.
  • it is interesting in which parts greek capitals, in an emphasized ductus, depart from the basic principle vertical-strong vs. horiz.-light. Have another look at a traditional Didone Greek and then explore how its rhythmical features could be adapted in your design.
  • Craig EliasonCraig Eliason Posts: 1,264
    edited December 2020
    Revised δσωΣΦΨΥ. (New versions on right.)
    I'd been concerned about /Phi and /Psi getting too wide, but I think that risk is worth taking for the solidity of the bar. (Lowercase versions will retain their single-pixel bar.)
    Happy to see that adding two pixels could do a lot to deLatinize the Upsilon. 
    Lowercase sigma still sucks.

    Edit: Obviously gets quite dark top right, but I think this sigma (right) is marginally better:

    Edit: And maybe this taller-bowled delta (right) works:

    Edit: This bigger bowled and simpler rho (right) is probably better. Is this beta also better?

  • Cool... Will you have an optical size axis to control the inter-pixel gap width? :p
  • Cool... Will you have an optical size axis to control the inter-pixel gap width? :p
    Well, I've got a lot of plans for what the pixel shape might look like. For right now I'm just trying to get the letter shapes solid. 
    Here's the current status of the Greek: 

    and here's the Cyrillic:

    (In both cases some of the last lowercase letters are invoked when italicized (with a slope not seen here.) 
  • Christian ThalmannChristian Thalmann Posts: 1,775
    edited December 2020
    You might want to add another pixel to the tonos to give it more solidity, right now it almost looks like the dieresis. 
    And should the cap tonos rise above cap height?

  • Yes, these diacritics are challenging given the coarse grid. Do you think the cap tonos should NOT rise above cap height? In looking up existing designs I found a split between those that didn't and those that did. Here (in line 4) is what the cap tonos looks like still rising above cap height but with an additional pixel below that tries to give it more length and angle. That line also shows the same alteration to the lowercase tonos, which is also what the Polish-localized acute looks like (line 2). Does it look too curvy? I could also try three pixels in a diagonal row, akin to the hungarumlaut in line 3. But that (and actually this curved version too) doesn't combine well for the dieresis-tonos. For context current acute/grave in line 1. 
  • Christian ThalmannChristian Thalmann Posts: 1,775
    edited December 2020
    I’m not Greek, but I like the cap tonos to align with the cap height. 
    For the lowercase tonos, I would suggest a two-pixel-high triangle:


    For the dieresis-tonos, you might get away with some merging?

    BTW: I’d suggest offering an alt for / that doesn’t duck its head, with a simple horizontal line as the breve. Otherwise it’s the only cap in Russian setting that doesn’t reach cap height. 
  • Adam JagoszAdam Jagosz Posts: 683
    edited December 2020
    That kreska is a nice try but looks quite a bit swashy. The standard acute you have would do better in this sort of font (or maybe better still, use what Christian suggested for tonos for kreska too). My inclination is that localizing kreska is text typeface territory anyway.

  • Here's the Thalmann triangle as tonos (incl. cap) and kreska. 
  • I am unsure about your goals:
    • Is the font to be used with actual low-res devices or is the pixelated look purely decorative?
    • What's the intended viewing distance: so small that individual pixels are discernible or so large that one gets the illusion of smooth strokes?
    • Who will use the Greek letters: Greeks for Greek text or scientists for formulas?
    • Is the Latin part of the font fixed or can it be changed as well?
  • Good questions! The pixelated concept is a stylistic choice; I'm envisioning this as a display face used large. Hope is that the type could be used as a display face in Greek- and Cyrillic-alphabet applications. Latin also still under development and in fact a number of letters have changed since the above posts. 
Sign In or Register to comment.