by erictom333
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.
116 Comments
Why does this font have ridiculous line spacing? I know it has some characters that go over the top and bottom lines, but I thought those determine line spacing. Ideally I'd like the lines to be ten pixels high, enough to make the box drawing characters line up.
I think it might be so the preview won't have glyphs that overlap vertically… You could try asking meek about that
Also, you missed a dot on the lowercase Roman numeral 12…
@erictom333 - Is it possible you have a stray brick floating high above one of your glyphs? If not, you can manually change the font height with a recent addition to FontStruct: Select Expert Mode, then Menu > View > Line Height. You can now adjust the blue horizontal line that indicates the upper boundary of the font's height.
@Goatmeal I've already checked for and removed any stray pixels, and changed the line height to nine pixels, exactly how I want it (per image). I'm okay with glyphs overlapping.
Things get weirder.
Top: Four ɧs. They should have one pixel of line spacing between them, as they do letter spacing.
Below: Box drawing and block element characters, which should line up.
Below that: The character with the lowest descender (a custom presentation form of ģ with the comma below) having zero spacing with a capital A; they should overlap.
Bottom: ÁáÉéÍíÐð. They all have their highest brick above the top line, and have been cut off. This does not happen with characters that go below the bottom line.
I'm not sure if this is intentional or not, but this glyph, where the red arrows are pointing at, to the right of a lowercase b resembles the glyph on the right half, and I'm unsure what languages use that…
@erictom333 - @BWM might be correct above; is the spacing issue you're seeing in the FontStruct preview? Is that where the samples are from? Or are you seeing it in other programs once you're installed the generated TTF? (In Notepad, WordPad, favorite word processor, favoprite image/picture software, etc.)
@BWM It's a handwritted variant of the letter b.
@Goatmeal The sample is from Notepad, not the FontStruct preview. Attached is a sample of text in Notepad in the font.
I'm currently debating where to put my own personal Latin characters. UCAS and Greek Extended are both good candidates, but I plan on encoding the correct characters there.
Pixel Sans now supports Devanagari. Example is from the Rigveda.
Bengali has now been added to Pixel Sans, as part of an ongoing project to include Indic scripts into this font. The following text showcases:
- This font's ridiculous line spacing (please ignore).
- The i and e vowel signs: unlike Devanagari's i vowel sign which must be placed before the consonant to display properly, Bengali's i reaches over the letter to display somewhat correctly. Similar with the e vowel sign.
What determines which fonts are worthy of staff picks? How can I improve mine so it does get a staff pick? Is it possible for the staff to re-review my font?
An insanely comprehensive character set. I salute your tireless dedication, enjoying especially the attention to detail from the mathematical operators to the miscellaneous character unicode blocks! And of course the range of languages present is impressive.
Can I recommend you add tags to detail the ranges you cover and make it easier for folx to find your great work here?
Also, pay special attention to the character width settings. It's easy at certain zoom levels to move it into apparent alignment with the grid when in fact is off by a hair. This affects pixel rendering – and I can tell you have at least a couple characters deep in the set with ever so slightly aberrant widths.
Overall awesome work, erictom333!
There is no strict criteria for Top Picks. It should generally be something that does not look similar to a Top Pick or something that has not been attempted before.
Very nice work, erictom333. Maybe when this is complete, it could be ported to OpenType and have alt glyphs in the .salt / .ss01 to .ss20 OpenType features (Placing them within .ss01 to .ss20 is recommended)…
@will.i.ૐ Thanks for the praise! For the latter, I might fix that later.
@BWM I won't usethe stylistic alt feature because I'll use this on programs like Notepad that don't support the feature.
There are two main features (or one feature and one bug) getting in the way of this font's development: line height and Unicode support. The former I've discussed above: every line should be ten blocks high, and any glyphs that go above or below the limit should not get cut off, but should overlap.
As for Unicode support: The font sorely lacks characters, mostly extended Latin letters, in a few blocks, including Latin-Extended E and the Private Use Area in the BMP and the tentative Latin Extended-F (10780-107BF) and -G (1DF00-1DFFF) (and the unallocated area of 1E600-1E7FF) in the SMP. (And a few other SMP blocks for historic scripts like Deseret and Adlam.)
@erictom333 FontStruct fails to save anything outside the BMP as far as I know, so I'm not sure about the SMP blocks…
The hangul (korean) characters don't work in this font.
I still love that it has so much compatibility!
I'm Not and Glyphs is Letter b is not good is not and good and so me Yeah is so Yeah is Not good is Not so good is not me yeah is Letter Glyphs is Good so Yeah Now you is not and what language use that...
Is Not Glyphs Letter Lowercase Roman is that numeral 12...
@Antoniolugb As I have already said, that's a handwritten variant of the lowercase b.
Pixel Sans now supports Bamum and Phags-Pa. As for the latter, I'm not sure if I should make the glyphs 5 pixels high (as I've done) or 7 pixels high; does anyone have any suggestions?
Does this look good, or would you prefer the Phags-Pa to be taller?
@erictom333: It is an impressive project and a huge undertaking. You must be an expert by now of vast stretches of Unicode land. I would increase the space a little. As William Leverette (will.i.ૐ) pointed out some of your glyphs are fuzzy/blurred and that is a mortal sin for a pixel font. Check out the highlighted areas. (The screenshots are truncated at 700 pixels on the left.)
@Frodo7 Thanks for the praise. My journey hss been a long one, my world map six sheets graph paper sticky taped together with every block of Unicode mapped onto it in a Hilbert fractal, my atlas the Unicode code charts. Also thankd for pointing out the spacing errors.
As part of Project Orisa, a project of mine involving adding support for India's nine official scripts, I present this sample, showing the seven scripts I have done so far and the two yet to go. And I'm not ending with the Malayalam chillu k, either.
Kannada and Devanagari extensions done. Also, 7000 glyphs!
Finished Malayalam, and by extension Project Orisa. But I'm not done yet.
As I approach 8000 glyphs (now including Sinhala, Ol Chiki, the rest of the Kangxi Radicals (some simplified), and Arabic presentation forms), may I request for the Private Use Area to be added? I have thousands of personal characters, far more than the High Private Use Surrogates block can hold, and I can't put them in another block because I might encode that later.
Pixel Sans now supports Canadian Aboriginal Syllabics (and a few more). Also over 8000 glyphs - 8677 in fact, beating the top Staff Pick (7:12 Serif) by 400 characters and putting Pixel Sans in the top 3 (behind only GS Unicode and no moral- [beta v1.7c], excluding password fonts).
Cool! I've been gone for a while for school and other hobby reasons. I think you're on track to having the most characters! Good luck!!!
Also, the only reason I have so many characters is because it's already possible to add characters in BMP blocks like the CJK, Hangeul, and private use region! If you know a bit of javascript, you can do some client-side magic to reveal these blocks (I think Fonstruct doesn't include these blocks because there are so many characters, it is normally beyond consideration to finish)
And besides, trying to load a very large block of letters can cause the tab or the browser to crash (depending on system RAM), so they'll have to be split up into smaller chunks of 256 or 512 if you do decide to add the CJK or hangul blocks…
With the forbidden power of hacking FontStruct, this font now has over 9000 glyphs. (Cue the obligatory meme.) Previously unavailable blocks added include Arabic Extended-A, Syraic Supplement, Latin Extended-E and the yet-unencoded Arabic Extended-B.
Cool!!!!!!
This font now has over 10000 glyphs (and the 3rd non-password font to do so!). Take these samples of the Metelko alphabet, Initial Teaching Alphabet, and Benjamin Franklin's phonetic alphabet, all of which this font supports but Unicode doesn't (yet).
neat!!! five digits gang! i'm only back until i have more glyphs than the password fonts, then i will likely be missing indefinitely again. this is a great way to fill the PUA!
How does Balinese look?
It looks pretty good, erictom333…
In the progress of encoding all 1164 Yi syllables; currently about a third of the way there.
A new milestone has been passed: After finishing off Lepcha, I have now encoded every Unicode character up to U+2000. I don't think any other font on this website has achieved this. Currently at 13,886 characters; first place is on the horizon.
Cool! I have finals and other projects to finish, so I'm adding nothing to my font for now. I'm sure you can have the most characters in less than a week!
Another hack I just discovered; Go to Inspect -> Sources, Ctrl+F "whitespace" (last result). There is a list of all the glyphs that can still save if you don't add any bricks but modify the width. Change the list to whitespace=[32,160,4448,4447,8192,8193,8194,8195,8196,8197,8198,8199,8200,8201,8202,8203,8204,8205,8206,8207,8232,8233,8234,8235,8236,8237,8238,8239,8287,8288,8289,8290,8291,8292,8293,8294,8295,8296,8297,8298,8299,8300,8301,8302,8303,12288]. Reload. You can now add more different width spaces.
? I wondered what that list did
Finally at #1 most glyphs. Changelog:
+ A few letters Germanic and Nordic dialectology that I forgot about
+ Vitaly Vetash's Interbet
+ 8-dot Braille
+ Some of the Symbols for Legacy Computing
* Moved Latin and Cyrillic alternate letterforms to the PUA
* Changed letters with cedilla
* Redesigned Mitrata Lipi
And I still haven't finished Yi yet!
That'll show those password fonts!!!
New milestone: 15k characters! Now including Old Italic, Gothic, Zelda symbols, more maths operators and symbols, more Coptic, and more!
Do you think the CJK Compatibility glyphs should be just like if they were normally typed (current), or as thin as I can possibly make them?
If they were normally typed (Same width as other CJK letters)
Issues:
-- Cyrillic
Cyrilic de (and derived letters) should descend, but doesn't. - https://en.wikipedia.org/wiki/De_(Cyrillic)
Same with Cyrillic tse. - https://en.wikipedia.org/wiki/Tse_(Cyrillic)
Same with Cyrillic dzhe. - https://en.wikipedia.org/wiki/Dzhe
Cyrillic uppercase u is malformed. - https://en.wikipedia.org/wiki/U_(Cyrillic)
-- Phonetic extensions
Upside-down æ (ᴂ) looks malformed. Remove the pixel in the top middle, and move it to the left middle. - https://en.wikipedia.org/wiki/%E1%B4%82
Sideways o (ᴑ) is missing a pixel. - https://en.wikipedia.org/wiki/%E1%B4%91
Rotated small cap R (ᴚ) is mirrored horizontally. - https://en.wikipedia.org/wiki/Rotated_letter
Superscript B (ᴮ) should be shrunk down to 4 px height to match the surrounding characters. Follow the example of the superscript E.
-- Letterlike symbols
Won (₩) should be wider to clearly show the horizontal strokes. Same with Naira (₦) and Ruble (₽).
TEL is set lower than SM and TM.
-- Number forms
Roman numeral 100,000 ↈ is squashed. It's allowed to write these glyphs like (((|))), so it's best to redesign the symbol and the related symbol for 50,000 (ↇ).
-- Enclosed alphanumerics
Parenthesized small letter i looks unbalanced; make the letter one pixel taller.
Parenthesized small letter t looks unbalanced; move the "t" one pixel up.
Circled capital letter O is missing a few pixels.
Keycap 10 is unrecognizable.
-- Braille patterns
The blank pattern should also take up the same amount of space as the other patterns.
-- Latin Extended-C
Subscript j should be set lower.
-- Coptic
Coptic alfa is incorrect.
Coptic capital sou should descend.
Coptic small ro's bowl should align to the baseline.
Coptic small psi should ascend.
-- Hiragana and Katakana
The letterforms are really amateurish, and nearly unusable for Japanese audiences, but that's to be expected. You could copy the letterforms from Misaki Font, an existing 8px Japanese pixel font.
Serious dedication, for real, especially for one person! I applaud you.
@Haley Wakamatsu Thanks for the praise and feedback. I'll adress each seperately.
- I've fixed tse and dzhe to have a descender, as per your suggestion and to keep consistency with other letters, like shcha and te-tse. Changed uppercase U and derived forms (including U+F23E) to be more like the lowercase U and Latin lowercase Y. I'm keeping de the way it is because I like it the way it is and because I'm not a big fan of capital letters having descenders.
- I'm not changing ᴂ, as it is exactly the same as æ turned 180 degrees
- ᴑ is an error that slipped past me; thanks for noticing.
- ʁ and ᴚ being the wrong way round is my error; I needed to distinguish the two for my own purposes. Fixed now.
- Fixed the letterlike symbols, number forms, enclosed alphanumerics, Latin subscript j, and most of the Coptic you mentioned. Kept Coptic capital sou as is for reasons I've already stated.
- I'll fix Braille blank later, I promise.
- Sorry about Hiragana and Katakana, they're not really my forte. Might fix them later.
Finally finished off Yi. Here's a sample. Aaaand the line spacing has gotten out of hand, could someone do something about it?
You already reached over 2^14 glyphs… That is some serious dedication over the span of 7 months… Excellent work!
This font has over 2.5x the number of glyphs as my largest font in terms of letter count…
WOAH! now thats a lot of tags! i even found this font by searching "Pixel"!
New feature: Devanagari half-forms, ligatures, and more abnormal pieces are now encoded in the surrogate blocks. Not finished. A sample will be uploaded later.
18k glyphs? That is really impressive, and yet, the download does not time out, which is a miracle, considering the number of letters…
And even with 12GB of RAM, my browser is struggling a bit to load (or update the contents of) the renderer…
@BWM Yeah, mine too. It takes forever to scroll.
While I'm here: This font stilll has ridiculous line spacing, even though I set a dedicated top and bettom. May I ask of the devs to fix this? I'd like there to be an override switch to set the default line spacing.
I have encoded a few "negative space" characters in the Specials block. However, when used they display as a very long space.
Finished off all 676 Minecraftian syllables. Here are five Minecraftian sentences (the latter four are pangrams).
A new milestone that has never before been achieved: This font now has over 20,000 glyphs.
Oh shoot you've got me beat now! Here's to another 10k glyphs? :)
I would also advise (although you may know this) that there's a technical limitation on .ttf fonts to only support a maximum of 65,535 characters. I don't know if you already have a plan for working around this - I'm splitting up my font (roughly) by plane - but I'll say so here just in case this is useful to you.
This font isstill not downloading correctly.
Added Marchen, used historically to write the Zhangzhung language in Tibet, but mostly because it looks cool.
Added Old Hungarian. However, it still doesn't download correctly; the font file displays an error when installed.
This font still does not install correctly. The following error shows up.
Should I split this font by plane?
I have, or at least I plan to. As .ttf and .otf fonts only allow for 65,535 characters, it's necessary if you want to eventually support all of Unicode (plus the PUA you have). I'm not sure what your end goal is (if you have one), but I may recommend it.
Okay, I might do that. However, I would greatly benefit from a "delete all" shortcut so I could delete all SMP characters from this font, and likewise.
@erictom333 By "delete" I wonder if you mean "delete the glyph completely" so it removes its entry in the font data so it won't be there in the downloaded font…
Yes,that's what I mean.
My computer is saying that it's not a valid font file.
I've been having that same problem.
It is definitely because there are glyphs assigned in the surrogate areas, which is causing the font to break (see sample)…
I hope Meek can get the glyphs outside the BMP to display correctly…
I don't think that's the problem. I couldn't downloadthe font ever since I added SMP characters, but I could before, even though it had characters in the surrogate blocks. I'll try to move all claracters out of the surrogates, though.
I've moved all glyphs out of the surrogates and it still doesn't download.
I will have a look. Thanks for the screenshots @BWM. Maybe the problem is specific to Windows. Anway, if there’s a technical issue like this, the best thing to do is to send a message via the contact page, describing what’s going on in as much detail as possible, with screenshots if relevant.
I'm pretty sure there are no glyphs left in the surrogate blocks, but it still doesn't download.
This font still doesn't download properly. Also, when I edit the supplementary private use planes, it lags horrendously.
From my contact with support, it seems like this is a problem with the SMP generally, since it also applies to the missing glyphs in the font window; it will be fixed but it may take some time.
Also - are you sure all of the glyph data is deleted from the SMP and surrogates? Even if you've deleted every pixel, the raw font data may still remain. I'm not entirely sure how to go about deleting it, but it may also be why in my Plane F font it shows 143 characters instead of 72, even though only 72 glyphs "show up".
I hope Meek fixes it, but in the meantime, if your font has SMP codepoints, you can generate the font through FontForge to get it to work with about 99.99% confidence…
This font is amazing. How did you make so many characters. I wish I could dowload it. And How did you made Bengali look good?. All other fonts of Bengali have that weird circule dot.
Actually, the dotted circle just indicates that a character is a combining diacritic.
And does Bengali spaeking people understand it?
How did you get the Devanagari to work?
Top and bottom diacritics have been encoded as zero-width characters. Left-facing diacritics (i vowel) have been changed to top diacritics, and a cloneof the left-facing diacritic has been placed in the PUA. Half-forms, stacked forms, and ligatures are in the PUA.
I'm sorry but your hiragana like ぎ and ぜ that have the characters connected to the accent look strange
For somereason, when I hover over a letter a popup does not appear with the letter name and hex code, like usual.
Progress on CJK ideographs is going agonizingly slowly, as the editor is extremely laggy. It's probably due to the large number of characters in the font and the large size of the block.
Heavily admire the dedication! I have no idea what I'll do when I get to here....
At this point, I'm starting to believe that this font will eventually have every character posible in unicode.
Good job!
@migyoroc The glyph limit for TrueType and OpenType fonts is 65,535 (including the .notdef), so that isn't really possible without creating multiple fonts (preferably with all of them having the same metrics as the original)…
@migyoroc Probably not. Not all of the CJK characters can fit into a 7x7 space.
Sorry, the kanji or hanzi should be one pixel higher then the Latin alphabet at least.
Im actually really impressed, keep trying to reach your goal!
This font wasn't really made with CJK in mind, that's why the 7x7 limit is so small.
Circle Sans
As a big miiddle finger to all the fonts trying to imitate mine, I've added Chakma and UCAS Ext-A.
You know what, that sounds like a good idea. Only qualm is I'd have to change all the other CJK-spaced glyphs to 9x9.
why dont you have some of the cjk filled in?
Great solution for capital ς!
@JoohnPlayz (JoohnFonts), sorry for offending you or wtv
Quite a few more clones of this font with barely or unchanged glyphs I notice.
Good Font!!
@erictom333 - How did you manage to do these unicode all?
a cursed number of glyphs
it got all unicode
@erictom333 I'd love you make a monospace version of this!
Well, looks like the font is corrupted. All the spacing are just incorrect. All the letters are VERY close to each other. I don't know why. I just download, and BOOM! Corrupted font.
Ok. Just found a limitation. It was downloaded as OTF, and it corrupted; it would be probably better if I downloaded it as TTF.
Well. I downloaded 'Creative Commons', and boom. It doesn't corrupt somehow.
Please add Mathematical Script Bold and Blackletter Bold
how did you manage to make 20k+ glyphs without going insane or losing your sanity
patience
Just noticed that the tally mark has 5 vertical marks and one going through all of them, which is 6 in total on U+1D378, when it should be 4 vertical marks, and one going through them.
how did you mess up that?
Please sign in to comment.