i have a similar question, is there a standard or any suggestions on measuring bitmap fonts? what should be the size of each square in fontographer, in order for the font to appear normal at 8, 10 or 12 pt? after a little research, i think that over at "04" they are using 125*125 squares and their fonts are appearing at normal size at 8pt i think... doesnt always work for me though..

As far as I know, 8 pixel fonts only show up as 8 points on a Macintosh monitor. On a PC monitor (VGA) they appear as 6 points, and on SVGA even smaller. To show up as 8 points, they need to be 11ppem for VGA and 13ppem for SVGA.

I've not done much work with bitmap fonts, but if you want to make an 8 pixel font, your "pixels" should be 125upm square. 10 pixel fonts need 100upm "pixles," etc.

> It's not easy to think up a satisfactory system of

It seems that, along with x-height, the cap height still plays a pretty big role in the visual size of a face, too. With lead type, we used to call that how big a face was "on the body." E.g., you could have 12 point type on a 14 point body. But it still has some effect digitally.

Looking at Bembo and Berthold Baskerville (not my favorite faces), Bembo's caps are 622 upms and Baskerville's are 706 - which seems a pretty big difference to me - while their x-heights are relatively equal (63% and 62% of the cap height, respectively).

It seems that this would have quite an effect on how large a face looked on the page.

>On the other hand, it does depend on the language...

This might be a bit off-topic, but one of the most interesting things I saw at TypeCon was Will Powers' presentation on book faces. Of the examples he showed, one was a bilingual (English and Dakota) book of stories. The English pages were set 10.5|13 and the Dakota was 10.65|13. I didn't get a chance to ask, but I'm assuming that's because Dakota uses fewer caps and extending letters, but has (many) more vowels. I never studied Dakota morphology specifically, but IIRC long- and short-vowels play a pretty big role in a lot of Native American languages.

Anyway, I just thought that was a really nice piece of type setting that did it's best to respect both languages, without creating a visually jarring spread.

deleted

> one of the most interesting things I saw at TypeCon

Yes, that's very interesting! Is Dakota a Native American language set using Latin? It might be more than the cap frequency that affects the relative size. Does Dakota use fewer extending lc letters?

hhp

Yes, Dakota has a Latin orthography - I'm not sure how/when it came about. My presumption was, just from looking at the pages, that the size is due to both cap frequency (much less than English) and lack of extending lc. Here's a sentence from Powers' book:

Mii dash ga-pi-izhi-giiweyaan, omaa dash indaa-ani-wiiji'izhinizhaawigoo owidi Mashkimodaang ezhinikaadeg, miinawaa go ji-gikinoo'amaagooyaan.

I might try to dig through some of my notes from Amy Dahlstrom this weekend. Most of her work was with Fox, but it might be interesting anyway.

Hi Nathan,
Thanks so much for your help.
Although I knew a bit about leads, history and the like, I wanted a more practical approach at the time of design, and that is what I get with your explanation.
Thanks again.
Regards,
Marcelo

Marcelo, the question may be basic, but it touches on important matters.
The whole concept of type measurement as we use it is archaic and
confusing. In the days of metal types, what made a 12pt font a 12pt font
was that the