Creating the new 7HR alphabet

It’s like a seven-segment display. But horizontal, so it’s on its side. But rotated, so it’s normal again. That’s what 7HR means—it’s 7-segment display horizontal rotated. So, the previous alphabet would be called 14N—it’s 14-segment display neutral. Seven segments is a lot less than 14 segments, so that’s already a great advantage of the new alphabet. In fact, I’d say it’s an upgrade across the board with only one minor drawback—that drawback will be explained later.

A horizontal seven-segment display (7H) works well for my alphabet.

Let’s looks at lines for place of articulation—place lines.

It has a labial line, representing articulation at the front of the mouth (i.e., lips).

It has a coronal line, representing articulation in the middle of the mouth.

It has a dorsal line, representing articulation at the back of the mouth.

Like 14N, all three lines are used to represent laryngeal articulation.

Now, let’s look at lines for manner of articulation—manner lines.

It has a nasal line, representing the position of the nasal cavity.

It has plosive lines, representing stopped airflow when used with a place line.

It has a fricative line, representing partially restricted airflow.

It has liquid lines…

…and approximant lines—both with partial lines reflecting partial turbulence.

It has transition lines, representing the transition with opposite partial lines.

To create a consonant letter, a place line and manner line are used together.

This is a labial plosive—i.e., «p».

Finally, let’s look at vowel lines. These only use four segments.

It has a close line.

It has a mid line.

It has a front line.

It has a back line.

It doesn’t have an open line or a central line, despite having an open central vowel—i.e., «a».  This is because «a» is so common and is often reduced, so I wanted it to be very minimal . So, the entire open central vowel is just one line.

Otherwise, to create a vowel, a close or mid line is used with a front or back line.

It could stop there, but it doesn’t. Unfortunately, 7H has two drawbacks: (1) when put into syllable blocks, «l» and «y» are hard to distinguish, and (2) it’s horizontal even though seven-segment displays are used vertically.

So, I rotated each letter 90° clockwise, which gives me 7HR. Of course, this has one major drawback: it distorts the featural system. There is still a one-to-one correspondence, but the labial line is now the top line…

…and so on and so forth.

Labial, coronal, dorsal, and laryngeal place lines.

Nasal, plosive, fricative, liquid, approximant, and transition manner lines.

Close, mid, front, and back lines, and the open central vowel.

Of course, just tilt your head back and all of the letters will still make sense. So, they are functional and I’d say they retain 90% of the featural symbolism. Again, the lines still have a one-to-one correspondence that maps out to the vocal tract, but it is a bit distorted.

So, here’s the labial plosive—i.e., «p».

The actual font is coming soon! I haven’t had the time to churn that out while moving into my new home (and enduring two and a half days of 80 °F heat because the capacitor on the HVAC condenser died). Plus, I also have a new new way of forming syllable blocks—the old new way was for 14N and was outdated by the new new way for 7H/7HR.

Greyfolk Font Update 3: Man, the Man Is Non-Stop!

Forget my previous post and even the one before it because here’s another update! I’ve been working on this font day and night like I’m running out of time. It happened after I woke up this morning after my update late last night. “There are just a couple of things I’d like to try,” I thought, then I’d go and finally play some Dishonored. Well, I haven’t played any Dishonored yet—I did this instead. What changed? The font isn’t as ugly as it was a day ago.

Yeah, that’s it. It only took me about 8 hours of trial and error. This program and that program and these settings and those settings and maybe I just gotta do this and maybe I just gotta do that and maybe it’s all because I made it monospace which introduced a non-breaking space that was messing up my syllable ligatures. Shh, it’s okay. I’m fine now.

Here’s the alphabet:

m n p t k f s h l w y a e i o u
m n p t k f s h l w y a e i o u

Here are some syllables:

pa  pla  pam  plam 
pa pla pam plam

Here’s a sentence:

pe  pom  te 
pe pom te
I love you

Here are some other (stand-in) characters:

c q _ ?
c q _ *

*stands in for any other character

Also, «c» and «q» don’t mean anything. «c» is my test vowel and «q» is my test consonant, and, together, they help me test syllable structures with «qc», «qqc», «qcq», and «qqcq».

qc  qqc  qcq  qqcq 
qc qqc qcq qqcq

My syllables are no longer utter trash. I rebuilt my font using BitFontMaker2 alongside FontForge because Glyphr Studio was being a butt. My biggest problem with Glyphr Studio is that it doesn’t respect the svg files that I import from Inkscape if I do so much as combine nodes and merge stuff (which I did to try to get rid of the problems I was having before like vertices disappearing and blurry glyphs).

I remade the basic font in BitFontMaker2, imported it into FontForge, created ligature glyphs in BitFontMaker2, imported those into FontForge, set up the ligatures through FontForge, and proceeded to pull my hair out until I realized that my syllable ligatures weren’t working (presumably) because I made it monospace in BitFontMaker2, introducing that pesky non-breaking space which didn’t want to be a part of the damn team and become a syllable ligature.

I’m gonna try to squeeze in an hour of Dishonored now and pray that I didn’t squeeze out the rest of my spine jelly. Really, though, I’m feeling much better, and I’m hoping to be powerlifting by the end of the week.

So, I’ve already updated the font for Greyfolk

It’s actually already reflected in my previous post. After I published that post, I realized that I was seeing some inconsistencies across my phone, my computer, and my girlfriend’s laptop. The font looked much sharper on my phone and on my girlfriend’s laptop while appearing fuzzy on my monitor. “How can that be?” I wondered. My monitor has a 4k resolution just like my girlfriend’s laptop, and I’d be surprised if my font showed up best on the mobile site on my phone (which it did). So, I went crazy. I knew that the font was not allowed to look better for other people when it wouldn’t even obey the display of its creator!

But why was it doing this? Does it have to do with anti-aliasing in Chrome? In Windows 10? Does it have to do with resolution or scaling or sub-pixel stuff? Well, no one really knows, unfortunately. Such secrets have apparently been banished from the internet. Luckily for my sanity, I’m a pretty good problem-solver.

When I first created the font, I was using 1024×1024 svg files with 64×64 squares to make a 16×16 grid to ensure that my font would be pixel perfect. Well, no, that’s not what happened at all. It was fuzzy because—for whatever frakking reason!—it was creating a bunch of sub-pixel colorful noise stuff around what should otherwise be pixel perfect lines.

So, I said yare yare daze and tried using 1200×1200 svg files with 75×75 squares to make a 16×16 grid in the hope that this would translate better to 12pt font. It didn’t. Then, I had the thought that I should create a smaller “pixel” within the squares. I did it this way and that way, but I ended up with 1024×1024 svg files with 64×64 squares to make a 16×16 grid except the “pixels” were (basically) 48×48 in the middle of the 64×64 squares, resulting in a “thinner” “pixel”. Yeah, I was definitely making stuff up, but it worked—for the most part. Like, 6 times out of 7, it’s much sharper than it was before, and, because I was already screwing around with the font anyway, I added a few new things.

Here’s the alphabet:

m n p t k f s h l w y a e i o u
m n p t k f s h l w y a e i o u

Here are some syllables:

pa  pla  pam  plam 
pa pla pam plam

Here’s a sentence:

pe  pom  te 
pe pom te
I love you

Here are some other (stand-in) characters:

c q _ ?
c q _ *

*stands in for any other character

Also, «c» and «q» don’t mean anything. «c» is my test vowel and «q» is my test consonant, and, together, they help me test syllable structures with «qc», «qqc», «qcq», and «qqcq».

qc  qqc  qcq  qqcq 
qc qqc qcq qqcq

These syllables are utter trash, but they illustrate the other end of my font problem. With the “thinner” “pixels”, some vertices just like to disappear. Seriously, just look at the difference of the «c» between «qc» and «qqc». In «qc», the «c» has all of its vertices, but, in «qqc», it doesn’t! At least, that’s how it’s appearing on my monitor.

For now, however, this is more than good enough.

Custom original font for the Greyfolk language

I’m pretty proud of this. Before my guest lecture about conlanging at IU on April 17th, one of my goals was to make a rudimentary custom font, specifically for Greyfolk GU. I’ve known for a while know what the letters look like and how they form into syllable blocks, but I hadn’t taken the time to turn that into a font. Well, I do now! It covers all of the letters as well as the ligatures for the «pa», «pam», «pla», «plam», «pe», «pom», and «te» syllables. So, let’s test it out.

m n p t k f s h l w y a e i o u
m n p t k f s h l w y a e i o u
pe  pom  te 
«pe» «pom» «te»
I love you

I hope that worked and that you can see the very beginning of what Greyfolk is supposed to look like! I took the liberty of making the syllables 36pt so they’re easy to read. I’m… not a graphic designer. Not even close. So, at a small size, they are pretty hard to make out.

pa  pla  pam  plam 
«pa» «pla» «pam» «plam»

This is still way better than I imagined I would be able to do on my own. There are just 693 more syllables to make! If each one takes me about a minute, I should be able to finish this font with 12 more hours of work! After that point, I think I’ll make it available to be downloaded.