I’ve been learning how to use FontForge and am now getting my feet wet with kerning. I see that there are two options to do kerning: classes and pairs.
To me, classes seem to have no downside. When I’m setting a ‘T’ to a ‘c’, ‘e’, and ‘o’, I can do them all at once! If I have something that’s a bit of an oddity, like an ‘a’ or an ‘s’, I can just set it by itself (knowing too that once I get into diacritics those characters won’t be alone anymore).
So, why would I use pairs at all? Is it just a legacy thing since pairs came before classes? Does setting up a class introduce a lot of overhead into the font file that can bloat it or slow it down?
pairs are generally more accurate than classes but take longer to create (obviously). classes are the most popular and the difference is negligible, but if you want perfection and a real “optical” look as oppose to a “mathematical” look, use pairs.
Source : Link , Question Author : Brendan , Answer Author : stephencosh