{ Multiple Master interpolation - where are the kerning pairs? }

Rene Verkaart's picture

Hi guys,

I'm finishing up a large font family and will make instances from my multiple master font of the weights inbetween my masters. This works just fine, but... it doesn't interpolate the kerning pairs. More than that, there are NO kerning pairs in my new MM font.
I've got about 900 kerning pairs in my masters. I did a class based kerning which looks just fine to me.
Peter Bruhn told me that the both masters should have exactly the same kerning pairs, otherwise no kerning is interpolated? Is this true?

Does anyone know why the kerning date isn't interpolated and how to correct this?

thnx,
®ené

blank's picture

I don’t know what is causing the problem, but I will say that not having to worry about stuff like this is why I use Superpolator.

Goran Soderstrom's picture

First, this has nothing to do with using Superpolator or not, or what do you exactly mean James?

Rene, do you have kern pairs in both (or more) masters that represents the same pairs?
Same amount of pairs also?

They must be the same if they are going to get interpolated, and they must represent same pairs also. Otherwise the odd pairs will not be interpolated.

When you say there are NO kerning pairs in your new MM font, what do you mean? Were you not taking out instances FROM an MM-font?

Rene Verkaart's picture

Goran, the pairs are exactly the same, I duplicated them. I just didn't create the same amount of kerning pairs, because the extra bold not always needed them.

Correction, there are 425 pairs out of 900, but NO kerning classes.
And yes, I created the instances from the MM font.

James, does Superpolator do the job correctly even if the kerning pairs and/or kerning classes don't match?

Goran Soderstrom's picture

Is it an upcoming Fountain font? In that case I think you should send the MM to Peter for inspection, these things are so much easier to problem-solve when one can have a look at the actual MM source file.

blank's picture

First, this has nothing to do with using Superpolator or not, or what do you exactly mean James?

In my (very limited) experience using MM in Fontlab some little things alway goes wrong. Fixing those little things can add up to a good deal of wasted time. As someone who still has a lot to learn about type design this is especially irritating because I really need to focus on problems with my type and not problems with my software. But I have never had problems with Superpolator as long as my paths and nodes match up, which is pretty easy to do with the compare script that Erik provides.

James, does Superpolator do the job correctly even if the kerning pairs and/or kerning classes don’t match?

Superpolator assumes all missing pairs to be zero, so values still get interpolated.

Rene Verkaart's picture

Goran: yes it will be a Fountain font. The contracts habe been signed ages ago, so I want to get this font into the world. I sent him the files before and he will have a look at them.

James: You last senctence is just what I needed to hear. I'm not that familiar with superpolator, but that's just the way it should be. No kerning pair means a zero value. So simple...

thnx

Goran Soderstrom's picture

No kerning pair means a zero value. So simple...

Yep, that's the way it should be. Hopefully in next FontLab update for those who dont use Superpolator :)

Rene Verkaart's picture

I've tried about every setting in Fontlab but on generating the MM font, the classes are just not there. I spent about 5 hours to match the kerning pairs in both fonts, but it makes no difference when I create the MM font. I have 290 pairs in both masters, but in the new MM font I just have 74!! It only took some of the UC pairs. What could have happened to the other pairs?
Does anyone have a thought on that?

.00's picture

nevermind

blank's picture

Maybe you should send the file to Fontlab and ask them to check it for problems.

Rene Verkaart's picture

I think I'll just buy Superpolator and do it right. ;-)

Goran Soderstrom's picture

Sorry, but you must be doing something wrong if it ends up like that, but you already got advice on what to do.

Rene Verkaart's picture

True. It's always difficult to communicate the problem if the person on the other end doesn't know about the prefs and settings I'm using. I just hoped someone else bumped into this problem before and might know an answer. Ah well...

Thanks anyway!!
®

Syndicate content Syndicate content