dpla's 3x3

by Christian Munk (CMunk)

Download disabled

The designer of this FontStruction has chosen not to make it available for download from this website by choosing an “All Rights Reserved" license.

Please respect their decision and desist from requesting license changes in the comments.

If you would like to use the FontStruction for a specific project, you may be able to contact the designer directly about obtaining a license.

3x3 pixel font with upper case, lower case, numbers, punctuation and no duplicates. Wait, some duplicates in punctuation: parentheses = angle citation marks and comma = low 9-shaped quotation mark.
Info: Created on 8th March 2013 . Last edited on 11th March 2013.
License Creative Commons
Fave Tags:
  • -


Good work! Quite creativee!
Comment by Noah (winty5) 8th March 2013
Well created G, N, m, n, u, v, w.
Comment by Account Moved (Groszak) 11th May 2013
Oh, it's not your renamed 3x3 font !

Cool, thank you !
(+ 1 to my collection…)

Since you named it with the abbreviations of my name
(is it temporarily ? - you won't fool Google in the end, LOL)
I think you are quite satisfied with it. 8-)
So, I'm going to check it !

Here's the preview with my ASCII sorting v13…
(it must support this basic code to be 'complete'.)
Comment by dpla 11th May 2013
{OK, FS crashes @ rescal. transl. img… I retry…}
Comment by dpla 11th May 2013
Now I can comment longer on these successive tests :
[PS. it took me quite a long time to explain several things…]



There are no duplicate as a digital, monospace font
Great ! BUT…
There are 3 duplicates when we use it in the real world.

Contrary to 'computers' (a word I extend to any related machine),
we, humans, need to write by hand, and remember things easily
We cannot store in our brain too much superfluous information ;
we think analogically, which relieves our quite flexible memory.
I made this choice and preference from the beginning, years ago :
A good micro font must be as easy as possible to be remembered.

This means designers or pixel artists have to avoid, as well,
and as much as possible, all the inconsistent positions,
that only a computer can store 'safely' (cf. my '*x*grid_safe')…
So you see where your 3x3 font is 'unsafe', now or again :
• “x” and “ii” are dupes for any human (you cheated* ! 8-))
• “"” and “''” idem
• “%” and “::” idem
Plus all (yes, all) your 1-px and 2-px wide chrs :
(i.e. the more blank columns [spaces], the unsafer for humans !)
• “i”, “.”, “:”, “'” and “|” : you ask us to remember 2 useless blanks
• “l” : you ask us to remember 1 useless blank, and 1 arbitrary right alignment
• “l”, “(”, “[” : you ask us to remember 1 useless blank, and 1 logical right alignment

► As monospace here, we, humans, cannot count the spacing safely
(believe me, we need to count the space chars too, eventually ;
to do so, it'd be simpler (OK, ugly too !) to left-align everything,
then we'd have to double check every chr in a list of widths,
in order to guess where such and such chr should be located
to get rid of any unsafe duplicate, e.g. in “i ii iii ; ix xx xxx”
which is quite difficult to 'decipher' :
before one can understand :
0100 0000 0100 0100 0000 0100 0100 0100 0000 ;
0000 0100 1010 0000 1010 1010 0000 1010 1010 1010
next :
010 000 010 010 000 010 010 010 000 ;
000 010 101 000 101 101 000 101 101 101
and finally (the dummy '.' as a space chr for this demo) :
i . i i . i i i . ;
. i x . x x . x x x
I hope that's clearer - at least it was or should, for any programmer).

► In monospace, when the chrs don't have an even width, it's messy
(and I guess you worry like me about the look of your small font ;
yes, 'messy', I know : it's the opposite with larger monospace fonts…
just because 1 empty colum = ⅓ chr, and 2 blank columns = ⅔ chr…)

► In monospace, you have virtually 512 possible glyphs with this 3x3 grid,
hence a max of 512 chrs, while only 400 possible glyphs can be imagined
when you take care of the user (that's what I mean by 'safe' glyphs),
i.e. when you don't 'cheat' to get almost 30 % more 'solutions'… :-)

► Human writing cannot afford monospacing with complicated blanks in it,
because I'd want my micro font to be popular (you think the same with yours),
which means it can't rely on a matrix per se : no square sheets, rulers etc.
Simplicity (I dislike the 'minimalism' wording) needn't crutches…
Besides, writing with tiny fonts in the real life does not require ink,
but any material that can behave like a pixel : lights, objects, persons…
It's the artist viewpoint, and it won't be less legible either. 8-)



Not exhaustively on purpose… (I won't unveil my own 3x3 right now ;-)

On part 1 of my ASCII sortin v13 (as see in the image above) :
• There are good and/or logical couples of chrs
• 'G' is uncommon (interesting, though many would read it as 'a') !
• 'N' too (many would read it as '~')
• '8' too (I would read it as a 'black u. p. triangle')
• '%' too (clever, though I would read it as e.g. '*')
• '!' too (a lot, but quite logical with the less uncommon '?')
• '2' and '5' look rather good (except for their weight, cf. '3'), …
• ',' looks illogical compared to ';'
• '*' is a great resort to the cropping trick (nice in your font)
• There are some perfect lowercase, you know ! (oh, shh !) 8-)
• E.g. 'f' looks great, while 'c' may look like 'w' or 'o')
• Most of your uppercase is even more accomplished. 8-)
• The weight (hence the style) can be improved here and there
• I definitely made a quite different choice in many of my chrs…

On part 2 of my ASCII sortin v13 (see by yourself) :
• Your '6' and '9' are very puzzling
• I think some LC are too large (all right, it's monospace ;-))
• Your 'G' is illegible
• So does your 'g' ('J' confusion)…

You obviously belong to the most skilled 'micro fonters', now ! 8-)
Many good findings, I applause !
I'm happy to see many solutions of my unreleased 3x3…
(which decreases the number of possible, endless discussions.)

I won't 'process' the great deal of non ASCII you provided here.
(It looks very kind all the same !)


I cannot rate it (and shall not, since I would have to do the same everywhere), but if I really were to

(i.e. a knife on my precious display), and only after you can make it safe for the humans (variable width),

I'd give 7/10, even 8/10 (mine being max 9/10 IMHO because it's - forever - imperfect about a few chrs). In

comparison, when usability matters, 4/10 is the maximum I'd rate for any imcomplete font (that are not even

ASCII, i.e. with 0 duplicate).


@Groszak : CMunk's 'm', 'n', 'u', 'v' and 'w' are the famous, classic (old) and IMO reliable ones, as seen

in the 8-bit machines in the 80s (and still in my mind)… History repeats !-)


Comment by dpla 11th May 2013
{Notepad + FS = destroyed text formatting, sorry…}
Comment by dpla 11th May 2013
I'm reading $ here as 8
Comment by Account Moved (Groszak) 12th May 2013
Erratum : I forgot to finish, and debug b.t.w., my § Duplicates, about the ease to remember in the real life :

• “i”, “.”, “:”, “'” and “|” : you ask us to remember 2 useless blanks ;

• “l” : you ask us to remember 1 useless blank, and 1 arbitrary right alignment ;

• “)”, “]”: you ask us to remember 1 useless blank, and 1 arbitrary, yet fairly logical left alignment ;

• “(”, “[” : you ask us to remember 1 useless blank, and 1 logical right alignment.


@Groszak : really ? Thank you, we all need alternate glyphs for this difficult character at this size, that's interesting !

I think the best solution for his 'quite creative' (I agree) 3x3-grid font is (again here) :
Comment by dpla 12th May 2013
@dpla Really. $=8 because have a circles that can be used as 8. Here is the correct example for $:
Because it's letter s with |.
Comment by Account Moved (Groszak) 12th May 2013
@Groszak :
You're right and logical ;
Myself, I saw the same,
but bearing in mind that '$" = 'S' + '|' it resulted naturally in a different glyph…
(Which induces a small conventional trick between some chrs I won't explain or unveil here…)
Comment by dpla 12th May 2013
oops, you are wrong glyph. $ is lowercase s with | mixed from this font:
Just to improve lower case i and j:
Comment by Account Moved (Groszak) 13th May 2013
oh, no! fontstruct bug: no bold stopping!
Comment by Account Moved (Groszak) 13th May 2013
oh, no! stopped? test
Comment by Account Moved (Groszak) 14th May 2013
Comment by Account Moved (Groszak) 14th May 2013
Here is correct examples for %:
__X X__
Comment by Account Moved (Groszak) 14th May 2013
Comment by Account Moved (Groszak) 14th May 2013
oh no, the (font)(font/) tag is not worked now!
Comment by Account Moved (Groszak) 14th May 2013

Pixel4x4: here

Comment by ImmaPooh 10th November 2016

@ImmaPooh: cool, I commented it too (just a few weeks ago).

@Groszak: lol, I experienced a similar bug lately in the parser (italic always on, and message persistant & undeletable… the "<" and ">" must not be used as chevrons, besides too big codepoints put an end to the text). Result: I deleted my "Odyssey2 extended" from FS (it's on my site currently).

@CMunk: I visited your website yesterday… saw a photo, an old blog… I think you cannot be everywhere (even Winty5 might be more a teenage electro musician at present(?)). So {I may repeat myself, sorry}:

• Your "dpla's 3x3" is of 2 x-height, all right, but the lowercase can be improved to look like my similar 3x3 font made at that time (still too many oddities IMO, but the effort is probably the greatest I saw at this size, with my advices and without them).

• Good uppercase.

• I have myself a few 3x3 numerics (with both x-heights); yours are fair (and logical).

•  Interesting 'symbols'. you added a great deal of (at least) 8-bit characters; unfortunately, this (extended) part cannot but be crippled with duplicates.

• You know like me that an actual text font must show a complete basic Ascii set without duplicates. Unfortunately, your font includes 3 possible dupes ('" :% ix when proportional), after you cheated to get more glyphs (almost unusable/invalid).

"dpla's 3x3" has thus a limited portability: it must be displayed monospaced, therefore this limits its primary use to a software environment (one needs to remember a table of offsets). On the contrary, a proportional 'micro' design (i.e. without any blank column in the glyphs) has the maximum portability (even to be turned into monospace), and the greatest choice of supports (for real life applications used by humans and without too many cheat sheets ;-)). You can wait more if you want to see my choices (imperfect, I know).

Comment by dpla 11th May 2017

Also of Interest


Get 10% off the world’s leading font editor for OSX.

More from the Gallery

Digital Romeby Christian Munk (CMunk)
Onelinerby Christian Munk (CMunk)
Runicby Christian Munk (CMunk)
Messing aroundby Christian Munk (CMunk)
Power Gridby Sketchbook B
Concentriumby Christian Munk (CMunk)
mnml blissby regular_one
Western Slabby BWM

From the Blog


Vertical Metrics, Improved Touch Support and More


Future Competition Results


Competition: Future


New Bricks: Square Connectors