Typophile

Script font metrics going wrong with contextual alternates in some apps

Tanya Davis

I am creating a script font with some slightly complicated contextual alternates.

It runs fine in the Fontlab quicktest and in Adobe applications but not in Word or Wordpad - that's as far as I've got with testing.

The problem seems to arise when substituting a contextual alternate where the alternate glyph has a different width from the default. I don't know if this is relevant, I really haven't a clue what's happening.

For example, in typing 'feel' there's a glyph e.af which is an e coming after an f. Typing this results in the first e being placed too far to the right so that there is a gap between the f and the e and the first e overlaps the second e.

I'm generating as .otf.

Any suggestions gratefully received, it's driving me nuts.

Thanks,
Tanya

Nick Curtis

This may be a dumb question, are the alternates kerned, either manually or by class, exactly like the letters they replace?

Tanya Davis

No, there's no kerning at all for lower case letters, the widths are different, though.

Nick Curtis

A picture is worth a thousand words; a screengrab of the letter combo in the FontLab Metrics window and in a Word document would be helpful.

Tanya Davis

letter combination with contextual alts above and without below in Fontlab and Word.

Nick Curtis

It certainly appears that Word is using the default width, rather than the alternate width. Someone far wiser than I may be able to tell you why and how to fix the problem other than creating ligatures instead of contextual alternates...

malcolm

Hi Tanya

If you have been testing and making amends using the same font name you may need to delete the Windows font cache file. You can find it in:
C:\WINDOWS\system32\FNTCACHE.DAT (for XP)

Do a restart straight after and a clean font cache will be created.

M

Tanya Davis

Thanks Oldnick, I think you're probably right and in the absence of anyone wiser I think I'll try editing the alternate glyps so they have the same width. I want too many alternates for ligatures to be a reasonable option.

Thanks Malcolm, that is very useful general advice.

Tanya

Captured from web.archive.org on 17 Jun 2011. view this capture