ctrlcctrlv / kjv1611

A complete digital OpenType font restoration of the typeface found in the 1611 King James Bible
SIL Open Font License 1.1
95 stars 6 forks source link

Only one dagger exists in the font, but the 1611 KJV contained many different typographical daggers #9

Open artistofmind opened 5 years ago

artistofmind commented 5 years ago

The cross symbol has a couple issues:

  1. It is too bold. The actual cross used in the Bible is much thinner.
  2. It needs to be superscript. Its top should align with the ascender of the lowercase "h," and its base should align with the beginning of the serif on that same character's left stem. (Cf. Gen. 8:7.) We might also consider adding a second cross character (U+271D rather than U+2020). This one should be approx. half the height. (Cf. Gen. 13:8.)
ctrlcctrlv commented 5 years ago

My cross comes from "cross w J.png". To me it looks correct.

cross j

Please provide images to prove otherwise. I cannot work with verses alone, sorry.

artistofmind commented 5 years ago

Do you no longer possess the facsimile?

ctrlcctrlv commented 5 years ago

I donated it, see https://github.com/ctrlcctrlv/kjv1611/issues/8#issuecomment-451638283

artistofmind commented 5 years ago

Oh! Well, that certainly changes things. Might your church let you borrow it, or bring your microscope to the premises? If not, I have both the facsimile (the smaller size) and the Furness scans. I'd be happy to send you the latter; are you aware of any free hosting service that can handle a 1.76 GB zip file? (Alternatively, I could mail you a DVD. I'd also be willing to send you my facsimile for as long as you need it, or I suppose I could scan specific pages from it. . . .) I have emailed you an assortment of pertinent images, and will make the habit of doing so as and when I notice anything else amiss. Please check your email; if there is anything you still need that I haven't provided, just let me know. I am at your service!

ctrlcctrlv commented 5 years ago

@artistofmind If I send you an FTP account via email, would you mind uploading the Furness scans to it? I have several servers under my control, no need to use a free service. :-)

ctrlcctrlv commented 5 years ago

And of course thank you for the smaller images you sent via e-mail already. I can see there are many kinds of crosses, not only the one that I chose to include as dagger (U+2020). I think I will use Stylistic Sets (ss01, ss02 ... ) to encode them in the font, you should be able to access them in CSS.

artistofmind commented 5 years ago

I have only found two different sizes of cross in the body text (the Garamond in the margins is another matter): the full-height and the half-height. All are aligned with the tops of ascenders (e.g. "h"), and all are the same weight, which is quite thin. None look like in your font. Not a single one I have found in literally thousands I have looked at. (I will let you know if I ever do find an exception. If & when that day comes, I will tell you before I do anything else, believe me!) I have a suggestion: since the crosses are all the same weight regardless of size (perhaps the printing press had a minimum point size), maybe make some optical versions that automatically replace † when the text size changes. (So one glyph, with optical weights.) This may indeed be the problem with your scan: perhaps it is not to scale. (I.e. a half-height version, which in your font you have blown up to full size, without adjusting the weight.) I also see the edges look very splotchy. I suspect this is an artefact of the conversion to only two colors when that facsimile was produced. To be more accurate, your drawing should probally follow the innermost parts of those "blobs," rather than the outermost. In particular, notice how much thinner the right-hand stroke is than the rest. I suspect an accurate drawing would make the whole cross that thickness. Notice also how you have "underdrawn" around the blobs at the edges of the J, yet "overdrawn" on all the edges of the cross. Hmm. Yes, I think even the cross in your sample is thinner than you have drawn it. :-/

ctrlcctrlv commented 5 years ago

I see what you mean. It's late here and I have errands tomorrow, but tomorrow afternoon I'll try to add some thinner daggers/crosses into the font. :-) Hopefully by then you'll have uploaded to the FTP server I sent you.

ctrlcctrlv commented 5 years ago

This should be fixed in sfnt revision 3.2. I'll close the issue when it's released

ctrlcctrlv commented 5 years ago

I've left this open because I'm still unsure as to how many daggers there are...sometimes they seem placed up high, other times low. I'm as yet unsure which dagger should be default and which alternate.

artistofmind commented 5 years ago

Can you give me an example of a low one? (P. / bk. / ch. / vs. №?) I have yet to find a single non-superscript cross, and I’ve cropped hundreds. As I said before, I’ve only found the two types: a normal one (which should definitely be default) and a half-height version of the same. But both are vertically aligned with the tops of ascenders, and both are as thin as anything gets on that printing press. Moreover, if there are outliers, they are almost certainly in the margins, not the body text. And as both the font & line spacing are different in the margins, that’s neither particularly surprising nor necessarily relevant. In my crops, I have focussed on crosses in the body; and these seem remarkably consistent in size, shape, weight & position.