How to import kerning from a font to an other?

André Simard

I work on a typeface, so far I’ve drawn the extreme font (ultra thin, light and Ultra) in roman and italic. I’ve also done the spacing, feature and kerning on the roman one. I know that I can import the feature from a font to an other and that works well. What I tried is to use my kerning from my light italic to become my basic one for my ultra thin and ultra. Can I do that, if so how?

Thanks in advance.

James Puckett

I usually just do this with Metrics Machine. It’s fast and reliable.

Kerning importing is simple in Fontlab, just export an AFM file and import it into the new one. Unfortunately it is buggy and kerns do not always copy. There is an old Typophile thread about this, but it is no longer accessible. IIRC Karsten Lucke wrote someone a script to do this, you could ask him about it.

And if you know Python it should be simple to do this with Robofab.

Jan Schmoeger

If I remember correctly, the kerning import in Fontlab will drop all zero values, so if you need to have pairs with 0 (e.g. exceptions from class kerns), use 1 instead of 0.

André Simard

Thanks for your answers. I had already done the James’s suggestion and nothing happens, seems, as you mentioned, the kern do not copy. I've also tried to copy kern from an other font from the opentype window, doesn't work either. I will wrote to Mr. Lucke

To Jan: Is there a Find/Change button to do this action?

Thanks

[Deleted User]

The user and all related content has been deleted.

André Simard

I probably did the wrong procedure. I will try it again. Thanks

Jan Schmoeger

André: Alas, I never found one, so has to be done by hand. But if you sort the table in the 'Metrics window' by value, it's fairly easy.

André Simard

I did'nt know that, great! Thanks Jan \·>

Rainer Erich Scheichelbauer

@paragraph Sorry for being off topic and a terrible nitpicker, but I cannot help it. In your screenshot, I see a kerning pair for T-napostrophe. I believe there is no point in doing that. Afrikaans ’n is a separate word, so it would only ever appear between spaces. But to make things worse, it is not even in use as a separate character. Instead, people just type an apostrophe and then an n. Besides, the corresponding Unicode code point (U+0149 LATIN SMALL LETTER N PRECEDED BY APOSTROPHE) is merely a ‘legacy compatibility character for ISO/IEC 6937’ (quote from Unicode Checker), and as such deprecated. One could use it as an AFR ligature in the locl feature, though.

Mark Simonson

I'm so happy to hear that. I annoys me that it's not called apostrophen (I did it backwards in a font once because of the way it's named). Now I think I will just forget about it in future fonts.

Jan Schmoeger

Thanks, Reiner. No more /napostrophe—hooray. Please keep nitpicking, any other beasts that can be safely ignored? I hate /louse ... should this be a different thread?

James Puckett

…should this be a different thread?

It should be a Wiki unto itself. This is something type designers have needed for a while—a relational database of every character and language so we can keep track of all this stuff. Something like the Diacritics Project, but in extreme detail.