I've looked at these for a couple of months now from very very close. Fonts are code and the nature of code is that it's never perfect.

It had to be done, as the Garamond I wanted just didn't exist. Started feeling a bit weird after three months working on variable cap height regular, bold, 2 italics, and bold italics like a madman, so I brought it to a close. Two weeks ago, I read how Jony Ive spent the last 4 years on his Baskerville with a group of type specialists, and I went: "Ha!" https://www.fastcompany.com/90888571/jony-ive-spent-the-last-4-years-perfecting-his-typeface-heres-why-hell-never-be-done

We always deal with interpretations of music, and we always deal with interpretations of typefaces (printed, processed on different screens). Pure form, whether it's a triangle or a Garamond a only exists in our mind. A lot of the early Garamonds were photocopying the shape of the metal, ignoring both the nature of print and the the nature of the screen. That's one reason why early digital typefaces were lacking soul. Studying Garamond's original prints, f.i., is like reading sheet music.

@reichenstein Early digital versions of “classics” were mostly based on master drawings for photo composition. These drawings were decades earlier usually based on prints of the largest sizes in metal. So they were already copies of copies. Adding insult to injury, early digital fonts were often digitised very hastily too in order to quickly fill the market gap.

@paulvanderlaan I see, I thought they simply transposed the shape in metal which would explain why most of them are so thin. I thought, yeah, "Metal is always thinner than the inked in letter and surrounding white light makes the letter look even thinner than the inked one". But being copies of copies and the hurry in which they were produced explains it even better. Thank you for the explanation!

@reichenstein @paulvanderlaan And some copied the relatively crude unit-isation from photo/metal as well.

@klim @reichenstein I’ve heard similar stories how the 18 unit system can be found back in some digital Monotype fonts but never found any examples.

@klim @reichenstein @paulvanderlaan Also some years ago, @BijouType mentioned which Monotype fonts still have the 18-unit system traceable in their digital versions, but I can’t find the email now 😭

@typeoff @klim @reichenstein @paulvanderlaan Monotype Modern and Monotype Grotesque are basically untouched, in regards to spacing

@typeoff @klim @reichenstein @paulvanderlaan There are probably others, if you poke around the other digitized versions of Monotype families first manufactured before the 1960s — especially the ones that never became too popular as digital families. Paul’s account of how digitizations were done is exactly how Robin Nicholas explained it to me, with one corollary: some of the less popular families had better conversions, because they were a lower priority and were less rushed.