Creating a reverse (white on black) font

Get help with FontCreator here. Please do not post feature requests or bug reports here.
Post Reply
adp-rk
Posts: 1
Joined: Mon Nov 14, 2005 11:46 pm

Creating a reverse (white on black) font

Post by adp-rk »

Using a black rectangle and reversing contour direction, I can edit an existing font to create a reverse white on black font. I would like to do this for all characters in a font, accounting for the highest character to ensure the "space" character prints sufficiently high to match the highest character with an accent.

Can this be accomplished in Font Creator without having to edit each character in a font individually (i.e. perform a global change to all characters in a font with only a few steps)?
Thanks.
Bhikkhu Pesala
Top Typographer
Top Typographer
Posts: 9873
Joined: Tue Oct 29, 2002 5:28 am
Location: Seven Kings, London UK
Contact:

Post by Bhikkhu Pesala »

Why do you need to do this in the font? It is easy to create this effect in any DTP program. Just select the text and make it white, and overlay it on a black background.

There is no one-step way to do it in Font Creator, though it is not especially difficult to paste a rectangle into each glyph. You might have problems if a word-processor added spacing between letters while justifying text.
My FontsReviews: MainTypeFont CreatorHelpFC15 + MT12.0 @ Win 10 64-bit build 19045.2486
William
Top Typographer
Top Typographer
Posts: 2038
Joined: Tue Sep 14, 2004 6:41 pm
Location: Worcestershire, England
Contact:

Post by William »

Bhikkhu Pesala wrote:You might have problems if a word-processor added spacing between letters while justifying text.
I tried producing such a font once. The font has lowercase plus capitals T and F and a full stop, so the font is far from complete.

However, I have just tried it with Word 97 on this Windows 98 PC and indeed there are problems with justified text, not only with the spaces between words but also with the second ti combination in the word justification.

The font is available from the following web page.

http://www.users.globalnet.co.uk/~ngo/library.htm

The font is http://www.users.globalnet.co.uk/~ngo/FFAEI009.TTF and is shown in use in http://www.users.globalnet.co.uk/~ngo/ffaep009.PDF where it is rather more successful than with the justified text experiment which I have just carried out.

I add that the font is not made with Font Creator but was made some time ago using Softy.

William Overington
Dick Pape
Top Typographer
Top Typographer
Posts: 1360
Joined: Wed Oct 01, 2003 1:19 pm
Location: North Dallas, Texas

More of the same

Post by Dick Pape »

It's perhaps late to put my 2 cents in here, but it has been my experience that if you turn justification off, ragged right, etc. an application, such as MS Word, will not put in extra spaces. You also have to adjust line spacing to fit the exact height of the letters between lines -- but it will all work if you have the font. The right font, of course, is one which also has the space glyph drawn as a black box.

There are a couple of tricks on glyph spacing so it has been my experience that drawing a full b-on-w font can get the sizes all corrected before you have to publish.
William
Top Typographer
Top Typographer
Posts: 2038
Joined: Tue Sep 14, 2004 6:41 pm
Location: Worcestershire, England
Contact:

Post by William »

> It's perhaps late to put my 2 cents in here, ....

It is always a pleasure to hear from you.

I just tried typing in Word 97 on a Windows 98 PC.

I was using 24 point and the font Font for an exhibition i009 which is available from http://www.users.globalnet.co.uk/~ngo/FFAEI009.TTF and began typing.

This is a test using a ragged right margin

I found that that gave a fine white vertical line between r and i in right, so I typed some more.

This is a test using a ragged right margin so as to observe what might happen and hopefully gain some insight and throw light on the situation.

Word 97 wordwrapped after right, might and insight and there was a fine white vertical line between r and i in right, before the first i of insight and between t and i in the second half of situation.

I copied and pasted into WordPad and the result was good with no such fine white vertical line at all, the text wordwrapping after right, what, gain and the.

Both tests do have a ragged right margin which looks strange, so it may need justification to look right in wordprocessing.

Part of my interest in producing white on black fonts is that they can be used in the Serif ImpactPlus 3D modelling program to produce an effect of incised lettering in stone or whatever. I like the idea of making a white on black font of cuneiform characters so that one can produce an impression of characters drawn on a clay tablet, though that would be more of an artistic idea than it looking exactly like an actual clay tablet with cuneiform characters drawn on it.

William Overington
Dick Pape
Top Typographer
Top Typographer
Posts: 1360
Joined: Wed Oct 01, 2003 1:19 pm
Location: North Dallas, Texas

Post by Dick Pape »

Hello William

I read your font and composed a simple a-z line in MS Word and also saw some white vertical lines - after the i or before the j, after the j or before the k and after the l and before the m.

Looking at your glyphs I found you were very consistent with each starting at 0 and extending the width of the letter.

I remember some conversation here on the Forum some time ago about mathematically calculating glyph widths in order to take into consideration grayscale problems or something -- it was too complicated for me to understand. It did suggest (AFAIR), a letter had to be some factor wide in order to achieve this mathematical perfection.

I think that's what we are seeing here -- the letters are not exactly the right size for the rasticizer (rasticiser) to place "cheek-to-jowl" - at every point size, so it rounds upward and goes forward a very tiny amount (IDNKWITA).

Therefore, some fonts I've dealt have extended each glyph -10 (up to -30) so as to overlap with the previous letter. Works every time, AFAIK. -10 may be more than what is needed but you get the idea.

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

Post by William »

Thank you for your comments.

> I remember some conversation here on the Forum some time ago about mathematically calculating glyph widths in order to take into consideration grayscale problems or something -- it was too complicated for me to understand. It did suggest (AFAIR), a letter had to be some factor wide in order to achieve this mathematical perfection.

It is possible that you are referring to my idea of mathematical pseudohinting. The idea is that at 12 point in Windows (at least on a Windows 98 PC, I do not know about other versions) unsmoothed black and white rasterizing is on a 128 font unit grid, at 24 point on a 64 font unit grid and at 18 point on a third of 256, which is mathematically 85.333 recurring though I do not know what Windows uses. Thus using points on a 256 font unit grid for a font that should be precise at 12 point, 18 point and 24 point. I tend to split 256 as 168 plus 88, with the idea being that at 18 point there is then approximately spot on alignment and that at 12 point and 24 point there is clear guidance as to whether to fill a grid square or not to fill a grid square: clear because, for example, a value of, say, one third of a grid cell black goes to no and a value of two thirds of a grid cell black goes to yes. I try not to split 256 font units as 128 plus 128 because although that would be precise at 12 point and 24 point, there could perhaps be confusion over what to do at 18 point, as in that a value of one half of a grid cell black could go to no or could go to yes. My Quest text font follows that reasoning. It thus looks, if I may say so, quite good at 12 point, 18 point and 24 point. However, it looks quite strange at 14 point!

> I think that's what we are seeing here -- the letters are not exactly the right size for the rasticizer (rasticiser) to place "cheek-to-jowl" - at every point size, so it rounds upward and goes forward a very tiny amount (IDNKWITA).

I opened the Font for an exhibition i009 font just before writing this post and found that i, j and l are all 344 font units wide, which is 256 plus 88 font units. So it looks like the rasterizer has perhaps rounded down the width of the clockwise contour (that is the black background) and rounded up the width of the glyph!

Most of my fonts use whole multiples of 256 font units as the width of each glyph so that problem would hopefully not occur, however, most of them have no white space at the left of the main contour so producing a white on black font is not an easy matter! Perhaps I can put a negative overlap on each of the glyphs? Maybe I place too much emphasis on this alignment of verticals with grid lines: it is a good technique yet it does get a bit restrictive at times!

William
Post Reply