Page 1 of 1

Fraction kerning

Posted: 19 May 2021
by SCarewe
Hello, upon reviewing KO's kerning, I stumbled across its deletion of my fraction kerning. Maybe my practice is misinformed, but so far, I have .numr and .dnom spaced regularly (auto-width from .sup components) and then would kern things like fraction four.dnom negatively, for example, just like seven.numr fraction, which also has a lot of negative kerning due to the diagonal of the seven. KO doesn't do this at all. Do I need to set up user-set auto pairs for all 1–9 numr/dnom against slash, plus models? Not that this is much of an effort, I was just wondering why KO doesn't take this into regard. Maybe a good addition to the auomated pair list.

Re: Fraction kerning

Posted: 19 May 2021
by Tim Ahrens
You are right, these autopairs should be generated automatically. It’s already on my list.

You will probably need to set up a couple of models, like one.numr-fraction and maybe fraction-two.dnom.

Re: Fraction kerning

Posted: 19 May 2021
by SCarewe
Yes, I set up those models already and it works great. Just wanted to suggest adding them to the autokern list, but if that's already on the list, perfect! Thanks.

Re: Fraction kerning

Posted: 16 Jun 2021
by SCarewe
Hello, sorry to revive this topic again. I have a question regarding the counting of "user-set autopairs". For some reason, when I manually set the autopairs for fractions, KO doesn't count them properly. Instead of having 20 user-set autopairs (*.numr|fraction, fraction|*.dnom), KO gives me something like 4, sometimes more, sometimes less. Even with subtracting the four to six model pairs I have set up for these combinations, this is far from my actual number. Yesterday, when I finished kerning with KO, I found that KO had simply disregarded most of my user-set autopairs and I had to re-enable them.

In the screenshot, for instance, fraction|four.dnom is set to auto by me, but doesn't appear in the list of four user-set autopairs.

Re: Fraction kerning

Posted: 19 Jun 2021
by Tim Ahrens
Thanks for pointing that out. I will have another look at the user-set autopairs and see whether it’s not storing them reliably.

Re: Fraction kerning

Posted: 16 Sep 2021
by Eben Sorkin
Maybe a workaround ( not ideal but workable if you have a tight deadline ) would be to decompose and manually space the numbers and slashes in fractions.

Re: Fraction kerning

Posted: 22 Sep 2021
by SCarewe
Hi again. I'm afraid I'm running into a lot of issues with small numbers and KO.

I have inferiors, superiors, dnom and numr. The latter three are components, automatically aligned, based off of inferiors. If I set them all up into the same special spacing group and set models in the inferiors, this doesn't translate correctly into the other small figures. If I go into the other small figures and set the same zero models, KO complains that some of these models contradict various other models and starts removing my capital kerning, or nn.

I suspect that KO does not correctly determine the different bounding boxes of the small figures, could that be the issue? Many thanks!

Re: Fraction kerning

Posted: 23 Sep 2021
by Tim Ahrens
Did you change the spacing of the inferiors while KO was running?

Which Kern On version are you using? The latest version (1.02) updates components if the base glyph is modified (whether contour or sidebearings). In older versions, this did not happen. If in doubt, close and re-open Kern On, and all the dependants should be updated correctly. Hope this helps!

Re: Fraction kerning

Posted: 23 Sep 2021
by SCarewe
Hello, yes, I did. I spent an entire evening trying to set up my small figures accordingly. What special spacing group do you advise? All into the same group? At some point, I was so frsutrated that I set up special spacing groups for each figure set, which also didn't work. KO 1.02, also restarting multiple times :)

Re: Fraction kerning

Posted: 23 Sep 2021
by Tim Ahrens
Sorry to hear you are struggling. Would you mind sending me the file?

Re: Fraction kerning

Posted: 23 Sep 2021
by SCarewe
Sent you my working file. Thank you very much for your time!