Calligraphy Experiment

Discuss FontCreator here, please do not post support requests, feature requests, or bug reports!
Post Reply
William
Top Typographer
Top Typographer
Posts: 2038
Joined: Tue Sep 14, 2004 6:41 pm
Location: Worcestershire, England
Contact:

Calligraphy Experiment

Post by William »

I have been trying an experiment.

Here is a transcript.

----

Tuesday 25 November 2008

09:19 am

Experiments with producing the effect of a calligraphic pen using FontCreator 5.6.

Start a new font, without predefined outlines, named Calligraphy Experiment 001.

Save As... CAEXP001.TTF.

Use Tools AutoNaming... to name the font Calligraphy Experiment 001 version 1.00 25 November 2008.

Format Settings... Metrics and adjust the metrics of the font to become 2048, -1024, 0.

Open the Stardisc font.

Copy the disc glyph and paste it into both the uppercase O and the lowercase o positions of the Calligraphy Experiment 001 font.

Scale the O by 75% vertically and 62.5% horizontally about the point (0, 0) without preserving side bearings.

Scale the o by 50% vertically and 37.5% horizontally about the point (0, 0) without preserving side bearings.

Consider the lowercase o.

Imagine that the thinnest stroke of the virtual calligraphic pen is 80 font units wide, so use Thin (40, 40) without preserving bearings followed by Hollow (80, 80) without preserving bearings so as to produce the effect of the letter as if draw entirely with a pen width of 80 font units. Keep a copy of the glyph in the p location while doing the above in case it does not do what I think it will do.

It worked as expected. Keep a copy of the resulting glyph in the q location.

The idea is now to copy the glyph to the r location, the r location just as workspace, and first delete the outer contour.

Now I want to rotate the contour clockwise by a particular angle, use Thin(80, 0) and then rotate counterclockwise by the same angle. Hopefully this will have the effect of having used a virtual calligraphic pen with the thin stroke at about the 11 o'clock and 5 o'clock positions and the thick stroke at about the 8 o'clock and 2 o'clock positions.

What should the angle be? Going back to the original design of o, it is as if in a box with sides in the ratio 4 high to 3 wide. A first candidate for the angle is arctan(3/4). Microsoft Calculator in View Scientific mode gives arctan(3/4) as 36.869897645844021296855612559093 degrees, so try 40 degrees. The capital O is as if in a box with sides in the ratio 6 high to 3 wide. The angle, if the O were considered alone, would be arctan(1/2) which is 26.565051177077989351572193720453 degrees, so whether to use the same 40 degrees for the capital O or to use a smaller angle is something to think about for later.

So, try to rotate the contour in the r cell by 40 degrees clockwise around the fixed point (0,0), thin using Thin(80, 0) without preserving side bearings and then rotate counterclockwise by 40 degrees around the fixed point (0, 0).

Delete the inner contour from o and paste the contour from r into the o.

Try the Fill outlines view.

The direction of the inner contour now needs to be changed.

All of the above manipulation of the inner contour was performed without saving the font between steps. Later saving may have moved the position of the glyph if anyone is studying this font later.

Try the same process with the O using the same 40 degree angle, thinking that a calligrapher would probably not change pen angle to draw capital letters. However, that is just a thought at the present, it is not based on any knowledge of calligraphic practice.

Validate the font.

Test the font using WordPad.

10:21 am

----

The font has letters O and o only at this stage, though p, q, r, P, Q and R have various contours in them as mentioned in the transcript.

There is no white space at either side of the O and the o. This will need to be added if the font gets developed.
CAEXP001.TTF
The experimental font
(5.25 KiB) Downloaded 445 times
Here is a graphic file made using the font at 72 point using Microsoft Paint.
The letters O and o at 72 point
The letters O and o at 72 point
CAEXP001.PNG (2.15 KiB) Viewed 3687 times
William Overington

25 November 2008
William
Top Typographer
Top Typographer
Posts: 2038
Joined: Tue Sep 14, 2004 6:41 pm
Location: Worcestershire, England
Contact:

Re: Calligraphy Experiment

Post by William »

Here is some more transcript.

----

10:49 am

Open CAEXP001.TTF and Save As... CAEXP002.TTF.

Use Tools AutoNaming... to make the font name Calligraphy Experiment 002.

Try a p by adapting a copy of o.

Add a vertical pale 127 font units wide. The value of 127 from the width of the o at the left side.

The coordinates of the pale are (0, 1280) (127, 1280) (127, -768) (0, -768).

Edit Select All

Edit Join Contours Union

Try a q by adapting a copy of o.

Add a vertical pale 127 font units wide. The value of 127 from the width of the o at the right side.

The coordinates of the pale are (641, 1280) (768, 1280) (768, -768) (641, -768).

Edit Select All

Edit Join Contours Union

Try a u by adapting a copy of q, along the way producing a first design for a as a by-product.

Format Settings... General and clear the one embedding flag that is set by default.

Try a b by adapting a copy of p, though do it so that a thorn is produced along the way as a by-product.
Or should a b not have that vertical piece down to the baseline?

Try an n by adapting a copy of p. Does that look wrong? Is it too heavy at the upper right?

Try a lowercase l by adapting a copy of b. Set the advance width to 128 font units.

Set space and nonmarkingreturn to have a width of 512 font units.

Try an h by adapting a copy of n.

Use autometrics to set 64 font units of white space at each side of the glyphs.

Validate the font.

Test the font using WordPad.

12:06

----
CAEXP002.TTF
(5.82 KiB) Downloaded 439 times
The display in the image below was produced in WordPad at 72 point then a Print Screen copy was made and pasted into Paint and then a section copied and pasted into Paint and then that image was trimmed and saved as a PNG.
Letters from the CAEXP002.TTF font
Letters from the CAEXP002.TTF font
CAEXP002.PNG (7.21 KiB) Viewed 3680 times
I noticed that the n might be too heavy at the upper right when making the font. Looking at the image the h seems to have the same problem though not quite so much and the u has the problem at the lower left.

Any ideas how I could change the design?

As I was making the font the o p and q seemed to remind me of an existing font but I could not think which font.

William Overington

25 November 2008
Post Reply