================================================================== The European Modern font set Copyright 2007 TeX Users Group. You may freely use, modify and/or distribute this file. ================================================================== * Computer Modern style fonts - * with accented and composite characters - * in ATM compatible Adobe Type 1 (PostScript) format - * useable with TeX as well as other applications. Background and Motivation: The European Modern (EM) font set consists of 57 text fonts that are vastly extended versions of Computer Modern (CM) text fonts. Each EM text font contains 310 glyphs compared to the 128 in CM text fonts (of which eleven are upright Greek letters). Each EM text font has over 90 accented characters, including the 58 `standard' ones found in most fonts in Type 1 format. * Having ready-made accented/composite glyphs in the font makes it * possible to achieve proper hyphenation when using TeX in languages * that use accented characters. This is not possible when using * Computer Modern (CM) fonts. Of course, proper hyphenation is already possible when using text fonts other than CM that have a suitable complement of accented characters, but the EM font set brings the `CM look.' And --- unlike the bitmapped EC/T1 fonts --- this font set is in scalable outline format. It also contains only 57 text fonts --- compared to the 565 fonts in the EC set. EM math fonts and CM math fonts: The EM font set covers text and math italic. This is augmented with CM math fonts for math symbols and math extension. Consequently, the EM font set also includes the CMEX*, CMSY* and CMBSY* fonts. For math italic one uses EMMI* and EMMIB* (in place of CMMI* and CMMIB*). The reason there are separate EM math italic fonts, is that CM math fonts do not contain upright upper case Greek. When using CM, the upright upper case Greek characters are obtained from the CM *text* fonts. In order to avoid having to use the CM text fonts, the eleven required characters have been imported into the EMMI* and EMMIB* math italic fonts. (While the EM text fonts do in fact contain the Greek letters, it is undesirable to have to compromise the font encoding in order to gain access to these glyphs). EM text fonts are just that: text fonts. They can be used with any encoding. The encoding of these fonte are not hard-wired like the CM or the EC fonts. Nor are these fonts limited to 256 glyphs (most contain 310). The glyph complement includes all characters in Cork (T1) encoding as well as in TeX 'n ANSI (LY1) encoding. Installation: The PFB and PFM files are supplied on four diskettes (or CD-ROM). Use Adobe Type Manager (ATM) to install the fonts. In Windows NT you can also install the fonts without ATM just by dragging the PFM files to the `fonts' folder. However, you will get better on-screen rendering if you do use ATM Deluxe 4.0 for Windows NT 4.0. TFM and AFM files may be found on the third and fourth diskette. TFM files for these fonts are not unique, since text fonts can be arbitrarily reencoded. We provide one set for TeX 'n ANSI (LY1) encoding (subdirectory texnansi) and one for Cork (T1) encoding (subdirectory tex256). TFM files for other encodings can be made using DVIWindo's `WriteTFM...', or by applying the AFMtoTFM utility to the AFM files. If you have Y&Y TeX System release 2.1 you will already have all the required TFM files (linked to appropriate encoding specific environment variables). If you have release 2.0, and you want to make sure you have the latest version, replace them with what is on the diskette. Typically this would be done as follows: xcopy a:\tfm\texnansi\*.tfm c:\yandy\fonts\tfm\texansi\tfm-em xcopy a:\tfm\tex256\*.tfm c:\yandy\fonts\tfm\tex256\tfm-em xcopy a:\tfm\*.tfm c:\yandy\fonts\tfm\nontext\tfm-emmi (Details may differ slightly depending on your directory structure). The first line copies the TFM files for TeX 'n ANSI (LY1) encoding. The second line copies the TFM files for Cork (T1) encoding. The third line copies the TFM files for the math italic fonts. Relationship to corresponding CM fonts: The EM font set has equivalents for 55 text fonts in the 75 font CM set. Note that there is no need for equivalents of the three CMTEX* fonts. This is because these are mere encoding variants of the CMTT* fonts, in the sense that glyphs with the same names are identical in shape in corresponding CMTEX* and CMTT* fonts. The only difference is that a CMTEX* font has a few glyphs not found in the corresponding CMTT* font and vice versa. The extra glyphs found in CMTEX* have been absorbed into the corresponding EMTT* fonts. They can be made accessible by reencoding these fonts, if desired. Similarly, CMINCH is just a different size of CMSSBX10 with fewer glyphs. Use EMSSBX10 at 109.5pt to get the equivalent of CMINCH (with more glyphs). (Where 109.5 = 72.27 * CapHeight / X-Height for this font). The reason there are 57 EM text fonts rather than 55 is that equivalents of two AMS text fonts (smaller sizes of the smallcaps face) are also included, for convenience. In addition, the math fonts supplied with the EM font set include the extra sizes of the CM math bold fonts found in the AMS font set, not in the CM set. This includes extra `in between' sizes not previously available in Type 1 format. Note that in the math fonts supplied here - both the math italic and the math symbol fonts - have been revised following Knuths recommendations. Plain TeX, LaTeX 2.09 Support: Support for the fonts under plain TeX, LaTeX 2.09 may be found on the third diskette in directories `plain' and `latex209'. There are files than can be \input on top of an existing CM-based format, as well as source files for constructing a new EM-based format. LaTeX 2e support: LaTeX 2e support for the fonts may be found in the `latex2e' subdirectory. Add: \usepackage[T1]{em} or \usepackage[LY1]{em} or \usepackage[LM1]{em} to your LaTeX source after \documentclass{...} and before \begin{document}. This will override the definitions for all the fonts and so the standard `fd' files are not invoked. You will also need to tell LaTeX what text font encoding you are using: \usepackage[T1]{fontenc} % for T1 / Cork encoding \usepackage[LY1]{fontenc} % for LY1 / TeX 'n ANSI encoding \usepackage[LM1]{fontenc} % for LM1 / Textures encoding If you add this line *before* the ones calling for the `em' package, then you can leave out the encoding option to that packages. Additional information may be found in the file `em.txt'. ATM problems with a few `non-standard' accented characters: Some versions of ATM have problems with accented characters not found in the Windows ANSI set. ATM 3.02 for Windows and ATM 4.0 for Windows 95 have such problems. ATM Deluxe 4.0 for Windows NT on the other hand works correctly. The problem shows up when the font is reencoded to expose accented characters that use `breve,' `dotaccent,' or `hungarumlaut.' These are accents that do not occur in the standard 58 accented characters found in Adobe fonts. The versions of ATM affected by this bug pick the character in the position of the accent in Adobe Standard Encoding (198, 199, 205) instead of the actual accent. If what appears at this character code is a simple character, it will be shown instead of the accent. If this character is a composite character itself, nothing will be shown. The work-around for this problem is to pass the font through the SAFESEAC utility supplied with the font set. This makes the font file larger, but avoids problems with these older versions of ATM. First convert the PFB file to PFA format, then use SAFESEAC, then convert back from PFA to PFB format. Using EM fonts with DVIPS: For use with DVIPS you will need to specify in the psfonts.map file what encoding you want to use. We provide two sample files, psfonts.emt for T1 (Cork) encoding, and psfonts.emy for LY1 (TeX 'n ANSI). The math fonts use fixed encoding and so do not need to be reencoded. T1 encoding contains accented/composite glyphs used in Latin writing systems of Western, Central and to some extent Eastern Europe. LY1 contains accented glyphs for Western Europe only, but covers many typographically important glyphs (like dagger and daggerdbl) for which there is no space in T1 encoding. The fonts can of course also be used with other encoding schemes such as Windows ANSI or Macintosh standard roman encoding. You also need to use the TFM files appropriate for the encoding you have chosen. We supply TFM files for the text fonts for T1 (in directory tex256) as well as LY1 (in directory texnansi). These TFM files are complete with ligature and kerning information. There is a single TFM file per font. * NOTE: No need for `virtual fonts,' or multiple TFM files! Special Features of EM when compared with CM: The glyphs in EM fonts that come directly from CM fonts have the same shape and size, and generally the same advance width and metrics as the corresponding glyphs in CM fonts. There are a few exceptions: (1) Proper subscript positioning with respect to upright upper case Greek letters: Since in the EM font set the upright upper case greek letters come from a math font (EMMI*), the appropriate modifications to the glyphs and the TFM file could be made to draw the subscript closer to the letter in the case of overhanging glyphs like Gamma, Upsilon, and Psi. This is not possible in CM fonts, because these letters are there pulled from a text font. (2) Hanging hyphens: The EM text fonts contain a second `hyphen' called `sfthyphen' which is the same shape and size as the regular `hyphen' but has half the width; as a result it `sticks out of the format' on the right. The `sfthyphen' appears at character code 127 in T1 encoding and 173 in LY1 encoding. Setting the \defaulthyphenchar to the appropriate value (as is done in em.sty) automatically generates hanging hyphens. If you prefer not to have hanging hyphens, just change the \defaulthyphenchar back to 45. (3) Minor corrections in some glyphs: In some EM text fonts, the letter `s' has slightly larger left side bearing, the letter `a' has slightly larger right side bearing, while the letter `t' has slightly less on the right than the corresponding glyphs in CM. (4) In some of the EM text fonts, the grave accent has been moved to the right and the acute accent to the left to get better appearance of ready-made accented letters. Also, the circumflex is a bit lower and a bit wider. The circumflex on `ecircumflex' has been moved a bit to the right. (5) Minor corrections in some metric files: Some of the EM text fonts have a negative kern for the pair `o j' --- unlike the corresponding CM text fonts which incorrectly have a positive kern. The italic digit `7' has a near zero depth --- unlike the corresponding CM text fonts where this glyph by mistake has a depth equal to that of the old-style digit `4' (which hangs way below the base line). (6) Kerning with respect to the `boundary' character: The quotation characters in the CM fonts include extra side-bearings to separate the quotation from the rest of the test. This assumes that the `English' opening and closing quotation convention is being used. This is a serious restrictions in some other languages where quotation marks are used differently. The extra side-bearing has been removed from the double quotation characters. Instead, suitable kern pairs with respect to the `space' character have been introduced. This works for quoteleft/right, quotedblleft/right, quotedblbase and quotereversed. (7) Kern pairs for accented characters: accented/composite characters inherit the kern values from the corresponding base characters (except in a few places where there are additional adjustments).