Kerning exception (lock icon)

I do not really understand the Lock icon functionality. What does it means that it is an exception, as I set the kerning by glyphs pairs?

In addition I have an issue with it: I’m working on negative kerning for these combos:
lbl lpl lql ldl.

When I change the left Kerning side of “p” (when after “l”) it changes the left Kerning side of “d” (when after “l”) too as they were linked.
I’ve tried to disable the lock but it appears again and again. It’s a real pain…

Then, sometimes the Kerning values are into parenthesis, sometimes not. Why?

Kerning exceptions only come into account if you have set up kerning classes - If you have anything entered into the “G:” (group) fields.

All Glyphs with the same value in the left or right “G:” field will have the same kerning value. Kerning is applied to the groups directly. But, and that are the exceptions, you can have kerning for the glyphs directly, too.

All “a” glyphs: a, adieresis, aacute … have the “a” as the value for both “G:”’s.
You add a kerning pair for “T” + “a” = -100. Internally the value is applied to “T” and “@MM_a” (the internal representation of a group kerning).
But then you want to have a different kern value for “T” + “ä” because the “dieresis” collide with the “T”. You add the kerning pair “T” + “ä” = -10. This is an exception to the class value and will show a open lock switch.

The parenthesis are only supposed to appear in the Metrics Fields. Are you sure you use the kerning fields (K: fields in the lower left and right corner) ? And do you have made the “d” from a component of the “p”? This would explain the strange linking…

I’ve used no components for these letters and no G values. And why it’s called “G”?If it is for “group” it’s unintuitive.

I really don’t understand… I add -10 between “l” and “q” and the App adds -10 BEFORE the “l” too…

And where I can change the G values for all associated glyphs?
If I do it in the box above the letter, I see no changes in other glyphs with the same class.

can you send me a screenshot of the window?

gosh. I abandoned kernings for these pairs. But I’ll do it the next time it happens.

By the way, what should the button next to “Copy” do?

In one Glyph I have a second layer (I even do not know why it is there), but I cannot delete it. I though this empty button would do something in this sense, but it doesn’t.

The empty button is supposed to delete extra layers. But it does not do so for now. Have to fix it.

I really don’t get it.

I’m trying to add a k+rounded class to “k” when it precedes letters as a d o q etc…

So i’ve created a text panel with all the combos : aka bkb ckc dkd etc till the end.

Now, I’m adding my k+rounded class to the first k in aka.
Suddenly all the k has these class! Is this the correct behaviour? If yes, I do not get its usefulness.
I imagine also these classes should appear in the Font info panel > Kerning, but mine is always empty.

Another case :
1.I’ve added my class to right G of k letter in aka.
2. the class is present in all letters combinations (sigh)
2. I add -30 in right K in k letter
3. I point on bkb; point on k and delete the class name.
4. -30 in aka disappeared with the class name…

I think you got the classes wrong.

Classes are there to have the same kerning values for several similar glyphs. So all glyphs with the same kerning group/class will kern equal.
e.g.: all the letters A, Adieresis, Aacute … get “A” as left and right kern class. If you now kern A+T, all other A… will be kerned the same.
You can also (depending on the design) assign “o” for all round letters.

o gets “o” for the left and right side
e gets “o” for left and e for right
p gets “o” for right
q gets “o” for left
odieresis gets “o” for the left and right side

If you now kern k+o, all letters with “o” on the left side will be kerned the same. This way you have all “round” letters kerned with the k.

Glyph classes are always letter based and the kerning is pair based. So you “tag” the glyphs, not the combinations.

Hop this makes it a bit more clear.

ok I see…
But I do not see the usefulness of it if I do not interact with glyphs. pairs I could simply define the glyph width rather than always kern it. I thought kerning is made to optimize spacing between specific glyphs couples.

Wouldn’t be more useful to define classes and then apply them to custom glyphs pairs?
This way I would recycle common kerning combinations when creating a new font and change the adjustment values only.

It is exactly what it is all about.

Look at this video
classes in FontLab
Don’t look at how you set up the classes - this is quite different, but on how to use them.

Thank you for the link. Now I’ve understood !
nice :slight_smile:

I’ve tested and understood the class method.
I only have an issue now : I applied “v” class to v and w. Now I’ve put -40 K K -40 to letter a. This way a kerns -40 both sieds with v and w.
This behaviour is not propagated to a+accents. Is it normal? Is there a quick fix to it or have I to add the value manually?

Do you have assigned the “a” class to the a+accents glyphs?

No. I though it’s automatically applied as the left and right margin.
I gonna do it! thank you