Artistic Text font

Gallery
1

I have now added the Artistic Text font to the web.

It is a font which I produced using the Softy program at various times during March and April 2004 and to which I returned yesterday in order to publish the font on the web. The published version was produced using Font Creator Program 4.5 from the Softy-produced original.

The files are as follows.

http://www.users.globalnet.co.uk/~ngo/ARTISTXT.TTF

http://www.users.globalnet.co.uk/~ngo/artistxt.txt

Both of the files are linked from the following web page.

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

William Overington

2 October 2004

2

Hello William

You have a very extensive catalog of work. Looks like a lot of thought and work went into it. Good results!

Some comments on Artistxt.ttf:

  1. Change the “not defined” char to something less bold. Overshadows real characters in char map. Dump two empty glyphs.
  2. Enable Postscript names. Add Macintosh mappings. Sort glyphs.
  3. Run Validate in FCP - resolve errors (5).
  4. Design preferences (IMO):
    a. ILTVZ and ijlz too conventional compared to rest of font. (Zz could have some of the 2 features. V could have v curves). T could be like palm tree (2 F. 1 reversed, without center bar). Not sure how to change IL.
    b. Æ æ Œ œ could be connected to each other (more ligature and less letters).
    c. 2 needs longer foot. 7 needs less angled stem. 4 and 5 horizontal arm too thin. * is too small to be legible.
    d. Letter space too wide.
    e. Design diacritical glyphs; build composite glyphs (not a problem now however). Felt accents were too far off lower case letters. I liked your å better than à.
    f. Some letters extend too far left (íîïì)and may collide with others. The sequence iïi gives 4 dots for 3 letters – kinda funny (rarely used grouping of course). Move the base (i) to right. R leg has unique bump in it - unusual for this design.

Dick

3

Thank you for your comments. It is nice to have feedback.

You have a very extensive catalog of work. Looks like a lot of thought and work went into it. Good results!

Thank you.

Some comments on Artistxt.ttf:

  1. Change the “not defined” char to something less bold. Overshadows real characters in char map. Dump two empty glyphs.

Regarding the two empty glyphs. The first was put there by the Softy program and the second, the one just before the space, was added by me quite deliberately using Font Creator Program 4.5 as part of the process of using the Softy-produced original to produce ARTISTXT.TTF.

Here is the section of the manual upon which I based the decision to do so.

quote

4.3.2 First Four Glyphs

TrueType outline fonts should have the following four glyphs at the beginning of a font. These were listed in Apple’s original TrueType specification. These glyphs are recommended to allow for the same version of the font to work on both Windows and Macintosh.

Glyph 0 is the .notdef (missing character) glyph.
Glyph 1 is the .null glyph; it has no contours and zero advance width.
Glyph 2 is the nonmarkingreturn glyph; it has no contours and positive advance width.
Glyph 3 is the space (and no-break space) glyph; it has no contours and positive advance width.
Glyph 2 and 3 should have the same advance width.

The .notdef glyph is very important for providing the user feedback that a glyph is not found in the font.
This glyph should not be left without an outline as the user will only see what looks like a space if a glyph is missing and not be aware of the active font’s limitation.

It is recommended that the shape of the .notdef glyph be either an empty rectangle, a rectangle with a
question mark inside of it, or a rectangle with an “X”. Creative shapes, like swirls or other symbols, may
not be recognized by users as indicating that a glyph is missing from the font and is not being
displayed at that location.

end quote

If I have got it wrong then I would be happy to change it, but I am under the impression from the manual that that is what I need to do.

However, having sought to use the contents of the manual as justification, I am aware that that section of the manual also makes recommendations about the .notdef glyph which I have not followed, leaving the .notdef glyph at my own design.

I recognize that the design might well be regarded as overshadowing real characters in char map, though the .notdef design which I use is designed so as to make a missing character very obvious in a block of text, though I do recognize that that reasoning is not perhaps as sound in relation to a display face as it may be for a body text face.

  1. Enable Postscript names. Add Macintosh mappings. Sort glyphs.

I tested the font using Serif PagePlus 9 on a Windows 98 PC so as to produce a pdf and checked in Adobe Acrobat that the font name came through, which it did.

Could you please elaborate on what needs doing with postscript names please.

In relation to Macintosh mappings I do not know what you mean. I use a Windows 98 PC. I have no experience of the workings of a Macintosh.

Does the font work on a Macintosh?

I tried sorting the glyphs on an earlier font and it put the two empty glyphs last. Maybe I should have highlighted all but them before the sorting. Anyway, I left them as they were quite deliberately after that as, thinking about it, I thought it good to preserve the order in which the glyphs were added.

I will try to have a look at the other matters which you mention and hopefully post again later.

William

4

Could you please elaborate on what needs doing with postscript names please.

I have now found section 4.5.5 of the manual and have been trying Format | Post… and the Include Postscript Names checkbox and the Generate Names button.

I notice that the .notdef glyph has the Postscript name quotereversed which seems strange, so perhaps there is an interesting history as to how that came about.

William

5

Hello William

After reviewing some of your references I am reluctant to speak a word, but online anonymity gives me “the strength…” to go on. (This has gotten too long and I’ve been unable to fully edit it, sorry. Gotta go to work).

  1. A more frequently used “not defined” character seems to be the empty box or rectangle. You did exactly what they asked – a character to clearly indicate “notdef”. It was a little too “clearly indicated” for me.

  2. I am PC only also so know nothing about True Type fonts on a Mac. I have found many fonts which properly assign characters based on the Microsoft or Unicode platforms and copy the same character numbers to the Macintosh Roman platform which is ok up to a point. Since they don’t always display the proper character name, “they should be corrected”. I have no idea whether this effort is needed, but it looks “nicer” when the name is the same as the glyph design. It does affect sorting sequence.

  3. I’ve applied this same rule to Postscript naming. While it’s not needed to use a font, it is very helpful (to me) to know how it has been mapped. Since FCP enables Postcript names and can Generate Names it is trivial to set up. You can quickly see some mis-maps because it’s got the Postscript name attached. I don’t know if there is another rule about needing PS names.

  4. This leads to the next point for which you showed the reference to the first 4 glyphs. I’ve read that also but not every font has them and in PC world it seems to not matter. It’s at this “no one will know” level I lose my moral way.

In your font the two extra glyphs are not mapped therefore are extraneous. They could have “dummy mappings” to assign them to those positions. Check other fonts for numbers to use.

The probable reason the notdef is assigned “quotereversed” comes from the “single high-reversed-9 quotatiom mark” mapping on the Microsoft Unicode platform. Just right click, select Properties and change it! Erwin has noted this condition (using the last mapping as the PS name) on other posts.

  1. Many features of true type fonts seem to try to integrate Mac needs to the font yet I have no way to test it (and not sure it matters!). I also have run into problems trying to get those 4 characters to the front of the font. I have tried various combinations of mapping and placement, but when I sort it doesn’t seem to work. Some times I give up.

I’m wondering if we have to follow those rules; if the Mac mapping is bad to begin with and if we don’t expect our fonts to run on the Mac platform, can we eliminate the Macintosh platform and mapping? The Sorting works nicely and consistetly (MS maps) if you do.

  1. I go through fonts and apply some things to all: proper platforms, complete mapping, having postscript names, Sort and Validate. Sort is especially helpful to arrange the glyphs in the same (albeit, apparently random) order every time. If there is a mismapped glyph it shows quickly after the Sort.

For instance, when I sorted artistxt.ttf the Eth, eth, Yacute, etc., sort to the front. Adding the Mac maps moves them to the end (where they should be). Getting rid of the Mac platform has the same affect. I don’t think you can “preserve the order in which the glyphs were added.” I think you’d rather know where to generally find Eth, eth, or Yacute.

Dick

6

Thank you for replying.

I think that a lot of the problems which I am having come down largely to the fact that I am trying to use the information in the Softy-produced fonts which I produced previously to try to produce fonts saved using Font Creator Program 4.5 rather than produce fonts entirely within Font Creator Program 4.5.

With the QUECHESS.TTF Quest Chess font I started with a new font in Font Creator Program 4.5 and also opened my Softy-produced original QUESTTXT.TTF font (not the present Font Creator Program 4.5-produced version of QUESTTXT.TTF, which had not been produced at that time) in Font Creator Program 4.5 and then individually found each desired glyph in the QUESTTXT.TTF font, copied it and pasted it into the new font. Thus the new font had all of the basic structure provided to a new font by Font Creator Program 4.5, including postscript names and mapping of the two glyphs which are unmapped in ARTISTXT.TTF.

So, of the six published fonts which I have produced using Font Creator Program 4.5, only QUECHESS.TTF has the mappings for those first two glyphs.

The fonts all work fine on this PC, running Windows 98, including the font name being displayed within Adobe Acrobat Reader when I have used the fonts to produce pdf files using Serif PagePlus 9.04. This matter of the font name being displayed within Adobe Acrobat Reader when I have used the fonts to produce pdf files using Serif PagePlus 9.04 was the main motivation for using Font Creator Program 4.5 to produce the published versions of fonts. Softy is a great program yet only has access to the Microsoft platform naming fields: fonts made with Softy always have New for the font name in Adobe Acrobat Reader. Well, that was the main motivation, yet I am learning more as I proceed. I am very impressed with the way that the Softy-produced files have been good for transferring into Font Creator Program 4.5. Apart from adding the extra glyph just before the space character in each converted font, all of the changes have been to naming fields: no changes have been needed to the glyph contours.

I remember reading once somewhere, probably on the Unicode mailing list, from someone at Microsoft who had responded to some question that the Microsoft operating system, when loading a font, uses some of the information from within the font file to produce tables which a font should properly supply directly, used simply out of necessity because so many fonts which are about had incorrect tables and that having the operating system produce proper tables was more or less a necessity if some fonts were going to be usable. It was before I started getting into font making. I do know that at one stage Quest text had a number of incorrectly-oriented contours and the Microsoft operating system tolerated them, so that, if, say, the outside contour of a glyph was counterclockwise it was treated as clockwise with no indication that anything was wrong. Also, I had produced an X as two overlapping rectangular contours and it tolerated that too. Well, someone kindly advised me that the X should be one contour and so I learned and my font was improved.

I do place great importance on precision and I like to try to do things properly, so although I could just have the fonts for a PC, I would like to try to produce the best quality product that I can so I shall try to learn how to make the fonts suitable for both a PC and a Mac.

Is there anyone who can try the fonts on a Mac please?

I have a policy of continuous improvement of my fonts, with the licensing such that once someone has purchased a licence then that licence is good for all future versions of that font, if indeed there are any future versions of that font.

William

7

Hello William …

I disagree with your statement:

a lot of the problems which I am having come down largely to the fact that I am trying to use the information in the Softy-produced fonts

I believe the important issue is your design!

What you are buying with a font is a theme or design. A purchase decision should be based on “is this font the best font for my project”.

A font may or may not have postscript names shown. It may have incorrect Macintosh or Microsoft/Unicode maps. Glyphs may have overlapping contours, duplicate points, single points, contours with incorrect direction (and the other 5 things FCP validates). Having those problems doesn’t necessarily affect the design, but only reflects on the execution. If you can make it install at all, it’s gonna be ok even with any execution faults because it is the best design for your needs!

As long as you have a correctly formed name (see Auto Naming) and have mapped glyphs someplace you can use the font in a project.

FCP helps so much that I believe some things should be correct in every font: character maps, validate errors, and a sorted font. As a last effort I might change a glyph to make it consistent with the font design. I’m bothered by fonts that don’t carry a theme all the way through (including the standard @ © € ® :trade_mark: characters). That is, if every other letter is italics they should all be italics or none of them… I throw away fonts if their design is bad; rarely if execution is bad. (I deal in free fonts, so it costs me nothing to be selective).

Use FCP to help you fix any execution problem it finds. But design is the key. We all have to focus on design. I look for good design, even though I don’t draw fonts, because I can fix bad execution. You have to develop a good design because it won’t be unique and noticed otherwise.

I refer back to my original comment:

  1. Design preferences (IMO):
    a. ILTVZ and ijlz too conventional compared to rest of font. (Zz could have some of the 2 features. V could have v curves). T could be like palm tree (2 F. 1 reversed, without center bar). Not sure how to change IL.
    b. Æ æ Œ œ could be connected to each other (more ligature and less letters).
    c. 2 needs longer foot. 7 needs less angled stem. 4 and 5 horizontal arm too thin. * is too small to be legible.
    d. Letter space too wide.
    e. Design diacritical glyphs; build composite glyphs (not a problem now however). Felt accents were too far off lower case letters. I liked your å better than à.
    f. Some letters extend too far left (íîïì)and may collide with others. The sequence iïi gives 4 dots for 3 letters – kinda funny (rarely used grouping of course). Move the base (i) to right. R leg has unique bump in it - unusual for this design.

These are all design consistency questions which should be addressed.

Good luck on your efforts to sell/license fonts. I have 5,000 free fonts that say it’s a very tough world out there.

Cruelly yours,

Dick

8

I have just updated the version of ARTISTXT.TTF which is available on the web to version 0.19.

This version has been produced today and has postscript names and 32 additional glyphs.

The font is available as follows.

http://www.users.globalnet.co.uk/~ngo/ARTISTXT.TTF

The font file is available from the following web page, yet readers who would like to have a look at the font before reading the note about a feature implemented in the new version of the font might like first to obtain the file from the link above.

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

This is, I feel, an interesting aspect of trying to become established as a supplier of fonts. This font now has a feature which a lot of display fonts do not have. Maybe the inclusion of that feature will make it popular amongst people who might like to use it. This is a feature which I am hoping to include in my fonts as a regular feature.

William

9

A new version of the Artistic Text font, namely version 0.21 has been available on the web since 6.58 pm British Summer Time (1758 GMT) today.

This version adds no glyphs. This morning I started a new font, Artistic Text 021 in ARTIS021.TTF, in Font Creator 5.0 and used the glyphs from the old version, copied one by one using the clipboard, as the artwork for the new font. I adjusted the metrics.

I then produced Artistic Text version 0.21 in ARTISTXT.TTF from ARTIS021.TTF. This involves, after using “save as” to produce the ARTISTXT.TTF file, deleting all of the unused glyphs in ARTISTXT.TTF, then using Format | Settings | Ranges and calculating both ranges, then validating the font. The font is then tested using Microsoft Word 97.

ARTIS021.TTF still has ready-mapped spaces for adding other characters. So any future development would start by copying ARTIS021.TTF to a new file ARTIS022.TTF and using the font name Artistic Text 022 during development. This means that I do not alter existing finished work, just in case later additions need to be rethought or abandoned.

The font is available from the following web page.

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

William

10

Artistic Text version 0.22 is now available on the web.

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

As well as now having the characters for Esperanto, Artistic Text now supports Latvian.

http://forum.high-logic.com:9080/t/latvian/1000/1

William

11

I have now updated the Artistic Text font to version 0.221 which differs from version 0.22 in that the flags for code pages Latin 2: Eastern Europe (1250) and Windows Baltic (1257) are set.

The flag for code page Latin 1 (1252) was set in version 0.22 and remains set in version 0.221.

http://www.users.globalnet.co.uk/~ngo/ARTISTXT.TTF

This change was because of a problem which is documented in the following thread.

http://forum.high-logic.com:9080/t/strange-problem-with-some-of-my-fonts-yet-not-with-others/2205/1

William Overington

11 September 2008