The Metrics View

The metrics view allows you to examine how characters look together and to alter the spacing between them.

If the metrics view is created from the font view then any selected characters (well, the first 15 or so) will be displayed in the view in the order they were selected. If it is created from an outline or bitmap character view then only that character is displayed. If it is created from a bitmap character view (or from the font view when that view is displaying a bitmap font) then the metrics view will display bitmaps, but will not allow you to edit the bitmaps metrics (you must do your metrics editing in a view displaying outlines).

Underneath the menu bar is a text field which allows you to enter additional unicode characters to be displayed.

Beside the text field is a pulldown list showing the script language combination that will be used for building kern pairs. (See the description on script language lists in the character information dialog)

If you wish to enter non-unicode characters you may drag them from the font view and drop them into the metrics view where you want them to go.

The mid portion of the view is taken up with the displayed characters. For outline characters these are sized so that the em-height of the font just fits in this region. There are grey lines between the characters marking the origin of the character on the right and the width of the character on the left. The baseline is also drawn in grey.

One of the characters is active. That character has a green line on its left (at its origin) and a blue line on its right (its width). You may change the width by dragging this line. You may change the left side bearing (lbearing) of the character by clicking on the character and dragging it left or right. You may change the kerning between this character and the one to its left by dragging the green line around.

The active character may be changed by clicking on any of the characters.

Underneath the display area are a set of text fields. You may type in new numbers for any of these fields to change the corresponding metrics of the character.

If the first character to be displayed is in Hebrew or Arabic (or any other right to left language) then the entire display area will be shown right to left

Note there is a very complicated algorithm given in the Unicode spec for displaying mixed right to left and left to right passages. This is not followed here. The direction determination is done by the first character alone.

Note also that neither Arabic ligature nor form (initial, terminal, medial, isolated) determination will be done in the display area (though they will be done in the text field above). If you want to see correct form you must enter the unicode character for that form (as opposed to the character for the letter), or use the View->Substitutions command.

You can make the window display vertical metrics and kerning by View->Vertical. (This is only available if your font has vertical metrics enabled, Element->Font Info->General->Has Vertical Metrics). In this mode you can change vertical width, vertical kerning, etc.

If you double click on one of the displayed characters then PfaEdit will bring up a outline character view to allow you to edit that character.

If the selected character contains active anchor points (that is something is attached to those anchor points) then you may move those anchor points around to see how they look. Note: When moving an anchor point in a mark, the result is counter-intuitive at first glance.

If you click in a character with the right button then a popup menu appears displaying all the simple positioning (GPOS type 1) features you have defined. You may select one of these and see the effect of applying it. Limitation PfaEdit currently only supports modifying the position, it will not display any modified advance widths.

Entering non-ASCII characters

Most of us are stuck with keyboards that only have ASCII characters on them. Or at best some sub-set of ISO 8859-1 characters. How can the full range of unicode characters be entered?

X supplies a sophisticated mechanism for entering CJK characters via stand-alone input method servers. PfaEdit supports this to some extent now.

Aside from that, many versions of X supply a "Compose Character" key, or something equivalent. The X software will map a sequence of keystrokes to one character in whatever the native encoding may be.

My keyboard doesn't have such a key, and my X doesn't do this. But I have implemented something along those lines for those of us with less expensive keyboards. If you are in a mode where text entry is possible then press the F12 key and follow it by a series of keystrokes. Accented characters may be built using the following:
     ` grave accent      0 ring      " double grave (or tonos)
' acute accent / slash . dot above
: diaeresis (umlaut) 7 breve , dot below
^ circumflex 6 caron 5 cedilla
~ tilde _ macron 4 ogonec

So  may be built by [F12] ^ A, and A with diaeresis and macron may be built with [F12] : _ A.

Greek letters may also be created. If you start with [F12] @ and follow with:
    A Alpha     B Beta     C Chi
D Delta E Epsilon F Phi
G Gamma H Eta I Iota
J (technical) Theta K Kappa L Lamda
M Mu N Nu O Omicron
P Pi Q Theta R Rho
S Sigma T Tau U Upsilon
V final sigma W Omega X Xi
Y Psi Z Zeta

This is the mapping used by the Symbol font. (Lower case letters are created similarly so [F12] @ a yields a lower case alpha). Accented greek letters may also be built up, [F12] @ " A yields Alpha tonos.

Finally the following special characters may be built:
    [F12] <space> <no break space>     [F12] * °     [F12] @ * <bullet>
[F12] @ <space> <em space> [F12] # £ [F12] @ # ¥
[F12] $ <euro> [F12] @ $ ¢ [F12] ! ¡
[F12] + ± [F12] - <soft-hyphen> [F12] @ - <en dash>
[F12] @ . · [F12] 7 2 ½ [F12] ; . . .
[F12] < <less or equal> [F12] > <greater or equal> [F12] @ > <triangle bullet>
[F12] = <quote dash> [F12] ? ¿ [F12] A Å
[F12] C Ç [F12] H <right index> [F12] O <OE lig>
[F12] P § [F12] [ <single right quote> [F12] ] <single right quote>
[F12] \ « [F12] @ \ <single guillemot> [F12] a å
[F12] c ç [F12] f <female sign> [F12] g ©
[F12] h <left index> [F12] m <male sign> [F12] o <oe lig>
[F12] p [F12] r ® [F12] s ß
[F12] t TM [F12] z <long s> [F12] { <left dbl quote>
[F12] } <right dbl quote> [F12] | » [F12] @ | <right single guillemot>

This still misses most unicode characters. But pressing [F12] twice will bring up a dialog which will allow you to select any character in unicode.

If you type [F12] by mistake then and Escape will get you out of accent mode.

Kerning By Class

Often it is a good idea to create classes of characters which kern alike. For example A, À, Á, Â, Â, Ã and Ä probably all kern alike (but note that o and ô probably kern quite differently after a T), and rather than create seperate kerning pairs for each of the "A"s above, they could all be placed in a class of characters which kern alike.

The Metrics->Kern By Class command provides an interface to this. It brings up a dialog showing the kerning classes currently defined for the font (it labels them according to the script in which they are active). You can create new classes, delete existing classes, or change existing classes.

Each kerning class needs a script/language combination for which it is active, and a set of standard otf flags for a feature.

But the more interesting stuff follows this. A kerning class actually consists of two sets of classes -- one set for the first character of the pair and one for the second character of the pair. Under each list of classes are buttons for adding new classes, and editing or deleting old ones. A glyph class consists of a space seperated list of glyph names.

Pressing [New] or [Edit] displays another dlg containing a text field with a list of glyph names. You may type in glyph names, or you may [Set] the glyph list to the names of the glyphs selected in the font view. Similarly the [Select] button will set the font view's selection to match the list of glyph names.

Class 0 is magic. You never set it, it contains any glyphs not mentioned elsewhere.

Underneath the classes is a display of kerning values. The possible first character classes are listed vertically, the possible second character classes are listed horizontally. When you select a class (in one of the class lists above) the display of kerning values will scroll to display that class. If you move the mouse into the display and let it rest, a popup window will appear showing the glyphs that make up the classes which intersect at that point. If you click on a kerning value in the display, the dlg changes to a kerning dialog.

The kerning dialog displays a representative glyph from each class and allows you to adjust the kerning between them (and all other class members of each class). You can select which glyphs to display in the pulldown menus. You can enter a new kerning offset by typing in a value, or you can click on the second character of the kern pair and drag it around.

-- Prev -- TOC -- Next --