Symmetry and /O

Michael Vokits

I've been trying to make better shapes for /O and /o and such, examining released fonts. I'd always just drawn a circle or ellipse, perfectly symmetrical left/right and top/bottom, and used that as a basis. Yet when I open up various forms of Palatino or pretty much everything with an angular stress, I see that it's more of a radial symmetry -- centering things at (0,0) never gives me an on-curve node at 0 (x or y), giving instead a more or less mirrored offset. The top point is always to the right of the bottom point, and the rightmost point is always higher than the leftmost. Here, I mean the outer curve, not the inner.



Looked like a pen tracing an ellipse. To test this, I drew circles in FontForge and expanded the strokes with a calligraphic pen, and always got the same consistent results -- exactly what I'd seen in the few fonts I have available to test.



I think I've been doing the wrong thing for years, and no wonder why my /O shapes are so bad -- highly symmetrical, yet simultaneously pointy and lumpy.

Anyway, the point of this n00bish missive is to ask if this is indeed best practice or at least normal (again, for the outer curves of angularly-stressed faces).

attar

I don't see why typefaces with angular stress would have a symmetrical O at all.

I'd practise calligraphy, it helps figuring things like this out globally.

Michael Vokits

Hi, Adrien. Indeed, although there is a radial symmetry due to tracing a symmetrical ellipse with a broadnib pen held at an angle, but that's different from a mirror symmetry. Spectrum seems to do the mirror symmetry thing in its outer curve -- its leftmost and rightmost extrema seem to be at the same height, and its topmost and bottommost extrema at the same length -- but that's judging from specimens, so I can't confirm.

Anyway, how does one get these shapes? Scans? Stroke expansion followed by fairly heavy simplification? Just nudging points on ellipses? The shape clearly derives from  simplifying a broadnib, so that's probably the best path to follow. (Pun unintended, but appreciated.)

Evie S.

For Itis, I chose a basic rotated form to provide stress, then I used the wonderful (fontlab/glyphs only, sorry) RMX Harmonizer went in.

Simply broadnibbing gives this result.

Albert_Jan_Pool

Calligraphy and angular stress: When you look at what the broad nib does, I’d agree. When you look at the pointed flexible nib though (Baskerville, Didot, Bodoni, Walbaum etc.), a vertical symmetry-axis surely makes more sense. When you take a look at the brush though, symmetry does not make that much sense at all. Note that Michael Vokits is right with his observation that Jan van Krimpen (Monotype Spectrum) and others did circular o-forms in broad nib / old style / Garalde oriented typeface designs. The same goes for Eric Gill (Gill Sans). In some of the original drawings of Jan van Krimpen you can see a tiny hole in the middle of the O because he used a compass to draw the outer outline. Also note that the use of classical / old style proportions rather provokes a circle oriented form for the o – this is in contrast to the classicist proportions as in Century Expanded, Univers, Helvetica etc.