Is there a reason for designing the regular weight between the ultralight and bold, I mean could you not design the regular and generate it with Superpolator using the ultralight and bold ?

I haven't really used MM to design fonts, but if I did I thought I should design the extremes and then interpolate the weights between.

Rr

Is there a reason for designing the regular weight between the ultralight and bold

Because the progression of changes in size and shape across weights is not linear. Then again, maybe it should be…

How functional linear interpolation will be, will depend a lot on the nature of the design, and how extreme your heaviest weight is. Is this serif or sans? In what general style? What kind of thick/thin contrast?

If you're using Superpolator, why build an MM *afterwards*? Or, if you're building an MM, why use Superpolator for interpolation first?

Note that I've never used Superpolator, just seen it demonstrated, so my questions may reflect my own ignorance of how it interacts with MM tech.

Cheers,

T

It’s a gothic sans, here’s a sample. If it looks a little screwy that’s because I haven’t generated new weights of light and bold since I changed n and it’s derivati:

As for jumping from Superpolator back to MM, I would need an MM font to use Remix, not because I actually want to do much MM interpolation. I could probably do the same stuff with Glyph Tweaker and Interpolated Nudge, but I think Remix would probably do a better job of simultaneous narrowing and thickening.

> but I think Remix would probably do a better job of simultaneous narrowing and thickening.

I think you're right on this. I use both Superpolator and Tim's Font Remix tools and have found the latter to be better suited for this task (narrowing/thickening). I think this may be one of its real strengths. As for your wider plan, I'm a little confused but I'm all for trying things!

As for your wider plan, I’m a little confused but I’m all for trying things!

I’m in the same boat—confused and trying things! I’m still very new to type design, so all this weirdness is, to some extent, an excuse to experiment with process.

I'd say, worry first about getting your g right. Making a superfamily is a pointless exercise in complexity until you've got A-z perfect. All you'll be doing is proliferating errors. Here you've got the classic problem of a glyph design that, light, just might print well at 8 points, but that, bold, will certainly clog. Do you redesign the character to something that interpolates more successfully, or do you change the design more radically, and uninterpolably, as weight progresses? Use a lot of paper!

Tom: "Is this serif or sans? In what general style? What kind of thick/thin contrast?"
Thanks, I needed that.

"It’s a gothic sans, here’s a sample."
Thanks, as I know it's hard to show in such an early state.

"...the progression of changes in size and shape across weights is not linear. Then again, maybe it should be…"
I do everything you list, to a subset of the complete glyph set, in FontLab. I think maybe you should let go of your feelings for tools and let the force be with you. Translated that means: Draw your thin, regular and fat, extra compressed thin, regular and fat, extra extended thin, regular and fat — until each of those nine works using whatever tools help your eyes. Then find the common point structure, if you have not already, then make more styles.

Cheers!

James, don't forget what normal MM inter/extrapolation in FL can do. Some of the things you describe can simply be done with FontLab's "Generate Instance" or "Assign Mask" commands.

Of course, Superpolator and Remix Tools have their special superpowers but for ordinary (global, isotropic) inter/extrapolation the results of MM/FontLab's tools will be identical.