First of all, letters with ink traps look like from one font, letters without - from second, and /I/ - from third different font.
Font with ink traps would be interesting, while the one without is quite generic.

Make all horizontals narrower for optical correction. Look at /T/ or /L/.
Get rid of /I/ serifs.
Try to apply ink traps whenever possible. For 90° connections, maybe cutting in at 45° angles would work.

Style unity in the typeface is the most important thing. What this one here doesn't have yet.

1) Total agreement on the ink traps. I've been thinking about adding them to the 90-degree connections and this convinces me.

2) I was going for ultra-minimal optical contrast between horizontals and verticals (it is there, believe it or not), but I guess this is "too" minimal. I'll bump that up (it was bugging me on the S anyway, so this is an excuse to fix it)

3) Unity was definitely a problem once I started writing sentences with it. The ink traps pretty much come out of nowhere, so it'll be nice to see it when it's more consistent.

This is exactly the kind of feedback I was hoping for and I should have another rev up before tonight. Thanks again!

The original post has been updated with a new PDF:

1) Ink traps are now a consistent part of the alphabet.
2) Horizontal strokes are now thinner.
3) As for the serifs on the I, I know that I'm fundamentally incorrect on this issue and always will be, but it's going to take a lot more shaming and criticism before I'll be convinced to remove it. I know it's wrong... but it just feels so right.
4) Lastly, a few character widths have been modified as well.

• Adding ink traps helped, but it's still not good enough. They differ in size compared to bigger, "normal" ink traps. Or try a variant with straight cuts, something like this:

• "Ultra-minimal" optical correction is the one where horizontals and verticals appear to have the same thickness. It's the required minimum. Forget math. What only matters is how it looks. (I know there was correction - I used a ruler).

• /I/ serifs have to be either removed OR made as thick as stems. Because current thickness aren't used anywhere (see #1 rule of typeface design above). It doesn't feel right at all, it's a stylistic alien.

• /I/ serifs have to be either removed OR made as thick as stems. Because current thickness aren't used anywhere (see #1 rule of typeface design above). It doesn't feel right at all, it's a stylistic alien.

You can always break ”rules”. And the “stylistic alien” of one can be the stylistic choice of others:
http://new.myfonts.com/fonts/bitstream/bell-gothic/black/

Doesn't look good there either. Besides, why make it SO similar to Bell Gothic?

I've finished the character set with the exception of glyphs based on others (such as diacritical variations). I've made a number of changes to the existing glyphs, added many new ones, and finally given up on the Bell Gothic issue. No more pseudo-serifs on the /I/. That one was just a matter of time.

Lastly, bear in mind that this font is still devoid of metrics and is currently relying solely on Illustrator's optical kerning. Each typeface has the same sidebearings of 35.

For some reason I can't upload files (HTTP Error 0 in upload/js), so I've linked them externally here:

http://www.alexvaranese.com/transfer/amv-antechamber_r3.pdf
http://www.alexvaranese.com/transfer/amv-antechamber_r3_flat.pdf
(I've uploaded the "_flat" variation since the font doesn't have hinting and isn't necessarily better when left as text in the PDF)

Sorry for the inconvenience, and if anyone can explain why uploads might not be working, that'd be great. I'm on the latest version of Windows Firefox and Flash.

Man, some chars are just awesome! For example 0, 4, X, all the punctuation.

Horizontal cuts didn't work out?

Alphabet would be nice so there's no need to hunt for individual letters.

Have you seen Retina?
http://www.moma.org/collection/object.php?object_id=139302

Thanks very much!

I tried horizontal cuts, and it had some benefits, but ultimately it looked less event (since the negative space was only being applied to one axis) and made consistency across the entire set of glyphs more difficult. With a few intuitive exceptions, the traps always come in at the halfway point of the angle of the intersection, which gives them a logical direction I can easily apply to any case.

Here are two new PDFs, one with a plain character set and another with a pangram. Still no metrics. And of course, this is a display face intended mostly for titling, but I consider a big block of body text (for which it's most certainly not intended) a good way to vet it.

www.alexvaranese.com/transfer/amv-antechamber_r4_glyphs.pdf
www.alexvaranese.com/transfer/amv-antechamber_r4_pangram.pdf

Well, some letters could have used horizontal and some - vertical cuts.

Also I was thinking - if you're going with display font, why not go all-in and make it stencil too. That way, /O, C etc./ would blend in better.

Did you mean adding a third cut that follows the arrow into the stroke from the counter? I played around with that after seeing your sketch and found it very cool but somewhat confusing. Maybe I missed something in your suggestion.

I also know what you mean about the stencil look; characters without any intersecting strokes definitely feel a little left out, but I feel that given the purpose of the typeface (a fairly standard condensed face with alternates for each glyph that will add a much heavier dose of stylization), I can accept the minor inconsistency for the sake of preventing the face as a whole from getting any more complicated.

I've finished the character set and given it what I consider "good enough" metrics and kerning. The spacing isn't going to win me any awards, but I think it's more or less on par with the typefaces that inspired it (Trade Gothic Bold, Univers Condensed Bold, etc.)

I'd like to get some feedback on how it seems to perform before I move on to releasing it for general consumption, so I've uploaded a beta version here:

http://www.alexvaranese.com/transfer/amv-antechamber_beta_zero.ttf

I've also added the lowercase set, which is actually a collection of alternate uppercase letters with heavy fills inside counters and between strokes. This particular feature ties directly into the video piece it's being created for, so it might make sense to ignore these entirely for the purpose of testing. The typeface is pretty conventional aside from the loweracse alts, and that's where I'm looking to vet it as thoroughly as possible. The alts are just a bonus for people like me that have specific use for them.

As a display face, I want it to be solid, reliable and usable, but it of course doesn't need all the nuanced perfection of a text face, especially since its intended application (titles in a video piece) tend to be hand-kerned a little anyway. But I'd like to make sure it doesn't strike anyone as a total mess.

The final version will be free, of course, so I'm not looking to "protect" anything here, but this beta has "BETA ZERO" written into the zero and letter O characters just to make sure anyone who helps me test this doesn't accidentally leave a potentially unfinished font installed.

Any feedback would be great. Thanks!

What I meant is instead of current /Q/ cuts, add that one.

Ink traps don't make sense unless you're planning on printing this face on newsprint, i.e. Bell Gothic, HFJ Retina. Ink traps in a digital face for screen use seem really bizarre.