Starting a new font - old regrets and mistakes

fontdesigner2's picture

What do you do when you start a new font? Anyone wanna name any regrets or mistakes they used to make? You know what I mean?

Té Rowan's picture

Ya forgot the nudge-nudge-wink-wink.

I'm yet to actually design a font (it's still just pencil drafts in a grid-ruled notebook lying somewhere about), but the idea came from reading about FE licence plate faces.

FE faces have the main design constraint that it must be difficult to modify letters to look like other letters in the face. (The Baader-Meinhof group used this tactic a lot.) I got thinking: Can this be done for US-style licence plates? Can it look good, too? And, can it be made such that a part can identify the whole?

Of speculations like that can a new font rise.

oldnick's picture

In another thread, someone suggested designing words, rather than simply letters. I can attest to the wisdom of this approach.

In one of the many lettering chapbooks from the 1900s through 1950s in my collection, I found a charming little alphabet with spritely upturned serifs, which I provisionally dubbed Twinkletoes. I constructed all of the characters necessary for a complete font, then exported them to Fontographer. It was only when I got to the kerning stage that I realized that virtually the only way the letters fit together without extensive kerning was the way I found them originally: that is, in strict alphabetical order. Bummer.

Chris Dean's picture

@fontdesigner2: What’s up with the angry slang shorthand? What exactly did you lose?

Thomas Phinney's picture

I think he lost his first draft of the posting.

One of the worst mistakes neophyte type designers make is to invest a lot of energy in kerning before they have really gotten the spacing quite right. The correct thing to do then is usually to throw away the kerning, which makes people very sad.



Ray Larabie's picture

Space the H, space the O. Make sure they look perfect, spend as long as it takes. Copy the H sidebearings to flat sides BDEFIKL etc and the O sidebearings to the round sides CDG. Fine tune the spacing but don't change H or O. I usually call those sides "solid" and I sometimes make notes so I don't mess with them. I use those letters to space the rest. For example, to space an S, use HSHOSOBSBCSCDSD etc. or make up words with those combos. After you're all done, you can fine tune everything and work on kerning.

It took me years to figure that one out.

Another tip: always select all/copy every few minutes when writing a post on Typophile. One wrong keystroke and it's gone.

fontdesigner2's picture

@Phinney - I remember reading you said that a long time ago. It's very true. The hard part is getting good enough at metrics that you can get them right before you kern. I still have to go back and forth and it takes me forever.

@Typodermic. That's great advice. I kind of figured that out on my own from opening up Helvetica or whatever font is similar to mine and looking at the sidebearings and using it as a rough guide. Then I made small adjustments until it fit my font, but that all makes sense. I'm gonna use this as a guide on my next font. And that select all copy thing is good advice too. Jeez I was angered by that! I had written this really great thing and I lost it all and I was so tired.

@oldnick. That's exactly what I did with this first font of mine. I would spell out words and tweak the letters until I liked the way it looked. Then I would make new words and make further changes. Do you ever find that you have to make compromises, where, in one word, something will make THAT word look better, but you find that you can't go that far with it? There were many times where I felt like I designed a letter to be halfway going in one direction, and halfway in another. I had to find a median, where it was safe, and consistent with the look and feel of all the other letters, but you can't JUST do that either. You have to form words and concentrate and making those words look as exquisite as possible. It's like walking a tight rope. It's hard to explain but I think you know what I mean. In the situation that you describe, it sounds like with a lot of hard work, you could could have fixed everything to work together. But the question is, is it worth it?

Thanks everyone

Syndicate content Syndicate content